WLOC and LMX

Unlike other leadership theories, LMX theory posits that the quality of relationships between supervisors and subordinates varies across individual subordinates. Accordingly, LMX research has highlighted the differential treatment that supervisors extend to in-group versus out-group members (Dienesch & Liden, Reference Dienesch and Liden1986; Scandura & Graen, Reference Scandura and Graen1984; Tangirala, Green & Ramanujam, Reference Tangirala, Green and Ramanujam2007; Wayne & Green, Reference Wayne and Green1993). In-group exchanges are characterized by high-quality relationships, while out-group exchanges reflect low-quality LMX. Subordinates in the in-group are typically favored by supervisors and receive greater access to resources, support, and developmental opportunities. In-group relationships are built on mutual trust, liking, and support (Martin et al., Reference Martin, Thomas, Charles, Epitropaki and McNamara2005). In contrast, out-group members tend to be disfavored, receiving fewer resources and limited interaction with their supervisors. Their relationships are often confined to formal employment contracts, lacking additional motivational efforts from supervisors.

While many studies have examined the antecedents of LMX, most have focused on the leader’s perspective, with relatively fewer exploring how subordinates’ characteristics influence LMX quality. However, because LMX is inherently a dyadic construct reflecting the relational exchange between leaders and subordinates. Thus, it is crucial to understand how subordinate traits shape their perceptions of LMX. Several studies explored how subordinates’ big five personality trait (Barrick & Mount, Reference Barrick and Mount1991; Bernerth, Armenakis, Feild, Giles & Jack Walker, Reference Bernerth, Armenakis, Feild, Giles and Jack Walker2007; Erdogan & Liden, Reference Erdogan and Liden2002; Graziano, Habashi, Sheese & Tobin, Reference Graziano, Habashi, Sheese and Tobin2007; Hogan, Reference Hogan1986; Perugini, Gallucci, Presaghi & Ercolani, Reference Perugini, Gallucci, Presaghi and Ercolani2003; Phillips & Bedeian, Reference Phillips and Bedeian1994), positive affectivity (Kinicki & Vecchio, Reference Kinicki and Vecchio1994; Phillips & Bedeian, Reference Phillips and Bedeian1994), and negative affectivity (Hochwarter, Reference Hochwarter2003; Hui, Law & Chen, Reference Hui, Law and Chen1999) influence LMX.

Drawing on social exchange theory (Korsgaard et al., Reference Korsgaard, Roberson and Rymph1998) that subordinates’ internal cognitive processes are central to how they interpret and evaluate relationship quality with their leaders, this study specifically focuses on subordinates’ WLOC as a key determinant of their perceived LMX. Locus of control (LOC) refers to an individual’s general and stable perception regarding the degree to which life events are controllable. According to LOC theory, individuals with an internal locus of control believe that life events result from their own actions and are therefore within their control. In contrast, individuals with an external locus of control attribute outcomes to external forces such as luck or fate, perceiving events as beyond their personal control. As a result, externally oriented individuals are more likely to view events as unmanageable and outside of their influence.

While LOC has been viewed as an individuals’ generalized trait, scholars have questioned whether individuals may exhibit distinct LOC tendencies in specific situations or contexts and explored context-specific LOC (Kalamas, Cleveland & Laroche, Reference Kalamas, Cleveland and Laroche2014; Wallston, Reference Wallston1992). Spector (Reference Spector1988) developed the concept of WLOC, which focuses on individuals’ perceived control over work-related outcomes. A meta-analysis by Wang, Liao, Xia, and Chang (Reference Wang, Liao, Xia and Chang2010) demonstrated that WLOC is a stronger predictor of work-related outcomes, such as job satisfaction, organizational commitment, and occupational burnout, than general LOC. Conversely, general LOC more effectively predicts broader psychological variables such as life satisfaction, affective commitment, and general burnout.

LOC is instrumental in explaining why employees with different personality traits may perceive the same organizational environment in divergent ways. For example, Judge, Erez, and Bono (Reference Judge, Erez and Bono1998) and Judge, Locke, Durham, and Kluger (Reference Judge, Locke, Durham and Kluger1998) found that internals generally exhibit more positive perceptions of the work environment, while externals tend to view it more negatively. Nystrom (Reference Nystrom1983) also found that internals tend to earn higher starting salaries and experience greater salary growth over time. Internals typically refrain from blaming external factors when they face negative situations. Instead, they take personal responsibility and actively seek to improve the situation. In contrast, externals encountering adverse organizational events, such as salary cuts or missed promotions, tend to attribute such outcomes to external sources, such as bad luck or unfair treatment by supervisors. Consequently, they are less likely to take proactive steps to improve their circumstances and are more prone to developing negative perceptions of the organizational climate.

Studies have found internals have positive relationship with or perception about leader. For example, O’Brien (Reference O’brien1984) reported that internal individuals tend to hold more favorable views of their supervisors compared to externals. Internals are more responsive to supervisory direction and show stronger commitment to their supervisors (Kinicki & Vecchio, Reference Kinicki and Vecchio1994). Lefcourt, Martin, Fick, and Saleh (Reference Lefcourt, Martin, Fick and Saleh1985) also found that internals tend to demonstrate better social skills and are more engaged in dyadic interactions, including active listening and socially adept behavior. Spector (Reference Spector1982) and Hawk (Reference Hawk1989) provide important evidence suggesting that LOC plays a role in shaping LMX. While most studies have examined LOC as a general personality trait, recent research such as Robert and Vandenberghe (Reference Robert and Vandenberghe2020) and Xiabao et al. (Reference Xiabao, Horsey, Song and Guo2022) have focused on domain-specific LOC, specifically examining WLOC within workplace settings.

Given that supervisors tailor their leadership styles to the characteristics of their subordinates, internals are more likely to receive supportive and effective leadership. Moreover, internals believe they can shape their work environment and interpersonal relationships, which enables them to develop stronger relationships with their supervisors (Martin et al., Reference Martin, Thomas, Charles, Epitropaki and McNamara2005). Kapoor, Ansari, and Shukla (Reference Kapoor, Ansari and Shukla1986) argued that internals are more likely to engage in upward influence tactics, actively shaping their relationships with supervisors in positive ways. Through adaptive coping strategies (Anderson, Reference Anderson1977), internals tend to build high-quality exchanges with their supervisors. They also demonstrate initiative-based behaviors, such as negotiating work assignments, seeking feedback, increasing communication, and using supervisors as role models (Dulebohn, Bommer, Liden, Brouer & Ferris, Reference Dulebohn, Bommer, Liden, Brouer and Ferris2012; Phillips & Bedeian, Reference Phillips and Bedeian1994). These behaviors often make them more favored by supervisors. Consistent with these observations, individuals with high internal locus of control also benefit more from mentoring relationships (Noe, Reference Noe1988; Turban & Dougherty, Reference Turban and Dougherty1994; Wang et al., Reference Wang, Liao, Xia and Chang2010).

Based on this, I propose that subordinates’ perceived control over organizational events plays a critical role in shaping the quality of LMX.

The mediating role of procedural and distributive justice

Subordinates’ perceptions of organizational justice are commonly categorized into two primary dimensions: distributive justice and procedural justice. Procedural justice concerns ‘feelings of fairness regarding the procedures used in an organization, including the processes for determining promotions, incentives, and dismissals’ (Leventhal, Reference Leventhal, Gergen, Greenberg and Willis1980; Thibaut & Walker, Reference Thibaut and Walker1975). Evaluations of procedural justice depend on whether organizational procedures are perceived as accurate, consistent, unbiased, and correctable (Leventhal, Reference Leventhal, Gergen, Greenberg and Willis1980).

On the other hand, Distributive justice refers to ‘feelings of fairness surrounding the allocation of organizational resources, including pay, bonuses, terminations, or any other resource that an organization can provide to employees’ (Adams, Reference Adams and Berkowitz1965; Roch & Shanock, Reference Roch and Shanock2006). This concept is rooted in Adams (Reference Adams and Berkowitz1965) equity theory, which posits that individuals assess fairness by evaluating the ratio of rewards to their inputs, comparing it to others or to their own expectations. Influenced by earlier theories such as Homans (Reference Homans1961) on reward-cost balance and Blau (Reference Blau1964) on expected returns, Adams emphasized that individuals care less about the absolute value of outcomes and more about their subjective perception of fairness in those outcomes. While other forms of justice – such as interactional justice (Bies & Moag, Reference Bies and Moag1986) and informational justice (Shapiro, Buttner & Barry, Reference Shapiro, Buttner and Barry1994) – have also been identified, the majority of the literature has focused on the two-factor model comprising procedural and distributive justice.

Procedural and distributive justice are conceptually distinct. As Cook and Hegtvedt (Reference Cook and Hegtvedt1983) observed, even when outcomes are distributed fairly, the perception of injustice can still arise if the procedures used are perceived as illegitimate or biased. Conversely, fair procedures may not offset negative perceptions if the outcomes themselves are seen as unjust. Therefore, examining both dimensions provides a more comprehensive understanding of how subordinates perceive fairness within an organization.

Prior research has examined the relationship between LOC or WLOC and LMX, consistently finding that individuals with an internal LOC tend to report higher-quality LMX relationships (Martin et al., Reference Martin, Thomas, Charles, Epitropaki and McNamara2005). Robert and Vandenberghe (Reference Robert and Vandenberghe2020) further extended this line of inquiry by demonstrating a positive association between domain-specific LOC (WLOC) and different dimensions of LMX. However, while these studies provide evidence for a direct effect of (W)LOC on LMX, the underlying processes through which WLOC influences subordinates’ perceptions of LMX have received limited attention, leaving the explanatory pathway largely unexplored. Given that the effects of distal traits are typically indirect, operating through more proximal cognitive processes (Kanfer, Reference Kanfer1990), it is important to explore the mechanisms that link subordinates’ traits to relational outcomes. Building on this idea, Scott et al. (Reference Scott, Colquitt and Zapata-Phelan2007) proposed a theoretical framework in which justice serves as a critical psychological mechanism that links subordinate traits to organizational outcomes by explaining the distal relationship.

Drawing from established findings on the relationship between LOC and justice (Miller & Nicols, Reference Miller and Nicols2008) as well as justice and LMX (Lind, Reference Lind2001; Tyler & Lind, Reference Tyler and Lind1992; Van den Bos, Reference Van den Bos2001), this study advances a more nuanced perspective by proposing perceived organizational justice as a key mediating mechanism that can explain the relationship between WLOC and LMX. The underlying assumption is that subordinates’ personalities (WLOC) influence the quality of their relationship with supervisors through their perceptions of fairness within the organization. More specifically, I argue that WLOC influences LMX not only directly, but also indirectly through its effect on perceived justice (both procedural and distributive).

Specifically, I argue that the positive relationship between WLOC and LMX is explained by higher perceptions of organizational justice among internals. Internally oriented subordinates are more likely to perceive organizational procedures and outcomes as fair because subordinates with an internal WLOC are more likely to feel that they can actively participate in or influence organizational procedures, which leads them to perceive these procedures as fair. High procedural justice enhances their engagement with the organization and their supervisor, resulting in better quality exchanges in the workplace. These interpretations are consistent with social exchange theory which proposes the quality of relationships in organizations is shaped by the fairness of interactions and resource distributions (Blau, Reference Blau1964; Cropanzano & Mitchell, Reference Cropanzano and Mitchell2005).

Similarly, subordinates with an internal WLOC tend to interpret outcomes as reflections of their own effort, reinforcing the belief that fairness is achieved through personal agency and thus, enhancing their perceived distributive justice. This inclination aligns with the just-world hypothesis (Lerner, Reference Lerner and Lerner1980), which suggests that individuals believe the world is fundamentally fair and that outcomes are deserved. When employees perceive higher levels of distributive justice (i.e., fair allocation of resources or outcomes), they tend to feel more positively about the social exchange with their supervisor, leading to improved relationship quality (LMX).

Given that both distributive and procedural justice are grounded in the fairness of outcomes and procedures within the organization, I propose that these two forms of justice mediate the relationship between employees’ internal WLOC and LMX. In other words, the impact of WLOC on LMX is channeled through the perceptions of fairness in how outcomes and procedures are handled within the organization.

The moderating role of cultural difference

Culture refers to shared values and norms that distinguish members of one group from those of another and often reflect general preferences for certain ways of thinking or behaving (Hofstede, Reference Hofstede2011). While various scholars have offered differing frameworks for classifying cultures (e.g., Douglas, Reference Douglas1973; Hall, Reference Hall1976; Kluckhohn, Reference Kluckhohn1962; Parsons et al., Reference Parsons, Shils, Allport, Kluckhohn, Murray, Sears, Sheldon, Stouffer and Tolman1951), Hofstede (Reference Hofstede1984) model remains one of the most widely applied, identifying five dimensions of cultural variation: individualism vs. collectivism, power distance, masculinity vs. femininity, uncertainty avoidance, and long- vs. short-term orientation. Among these, this study focuses on the individualism–collectivism (I–C) dimension, as it represents a foundational and widely validated distinction between the United States and South Korea. This dimension is particularly relevant for understanding interpersonal dynamics such as justice perceptions and LMX, as it captures how cultural values shape individuals’ expectations for fairness, autonomy, and group belonging.

The I–C dimension reflects the degree to which individuals are integrated into groups and the nature of relationships within a society. In individualistic cultures, such as the United States, people are expected to look after themselves and prioritize personal autonomy, independence, and self-expression. Social ties tend to be loose, and individuals are encouraged to speak their minds and pursue personal goals. The self is typically defined in terms of ‘I’ rather than ‘we’, and one’s identity is shaped more by individual traits and achievements than by group affiliations. In contrast, collectivistic cultures, such as South Korea, emphasize strong, cohesive in-groups, such as families, clans, or work units, which provide protection in exchange for loyalty. In these societies, individuals are socialized to value group harmony, conformity, and interdependence. The self is defined relationally, and people are more likely to make decisions that preserve group consensus and reflect shared obligations. As a result, behaviors such as self-restraint, maintaining face, and avoiding direct conflict are more common in collectivist contexts (Hofstede, Reference Hofstede2011; Singelis, Triandis, Bhawuk & Gelfand, Reference Singelis, Triandis, Bhawuk and Gelfand1995).

The cultural differences shape how individuals develop interpersonal relationships (Gudykunst, Reference Gudykunst1989). Collectivists tend to define their expectations of the organization through a relational lens (Erdogan & Liden, Reference Erdogan and Liden2006) and are more likely to tolerate unfairness when it serves group cohesion. They are often attuned to how decisions affect the collective and interpret fairness in terms of relational harmony and social obligations. Erdogan and Liden (Reference Erdogan and Liden2006) found that collectivists are less influenced by justice perceptions when evaluating their managers. For collectivists, maintaining harmony and fulfilling role-based duties take precedence over asserting fairness concerns (Markus & Kitayama, Reference Markus, Kitayama, Strauss and Goethals1991; Wasti, Reference Wasti2003). As a result, procedural and distributive justice may have a weaker influence on LMX development in collectivist contexts. In contrast, individualists place greater emphasis on fairness and are more sensitive to injustice, particularly when it directly impacts them. They are less tolerant of unfair procedures or outcome distributions and more likely to evaluate leader relationships based on personal benefit and equity (Ohbuchi, Fukushima & Tedeschi, Reference Ohbuchi, Fukushima and Tedeschi1999).

Justice perceptions vary depending on individuals’ cultural values (Pillai, Scandura & Williams, Reference Pillai, Scandura and Williams1999). Mueller and Wynn (Reference Mueller and Wynn2000) found that the importance placed on justice was highest in the United States, with South Korea ranking lower. Individualists are especially sensitive to procedural and distributive justice, as fairness reflects their personal rights and how they are treated (Earley & Gibson, Reference Earley and Gibson1998). For them, justice is closely tied to individual freedoms. In contrast, collectivists are less likely to form justice perceptions based on quid pro quo reasoning and place less emphasis on individual entitlements (Doney, Cannon & Mullen, Reference Doney, Cannon and Mullen1998).

Based on this, I expect cultural context to moderate the indirect relationship between WLOC and LMX via justice perceptions, such that the mediating role of justice is stronger in the U.S. context. Figure 1 presents the conceptual model of this study.

Figure 1. Conceptual model of WLOC, justice, LMX, and culture.

Sample and procedure

The data for this study were collected from both the United States and South Korea. The U.S. data were gathered from undergraduate and graduate students at large public universities in the northeastern United States. Participants were asked to answer questions regarding their current or most recent job. Participation in the survey was voluntary, and undergraduate students received extra course credit for their involvement. Prior to completing the survey, participants were informed about the study’s purpose, as well as confidentiality procedures. They were also advised that their participation was voluntary, and they could withdraw from the survey at any time.

A total of 316 participants completed the U.S. survey. Since sufficient time is required to develop justice perceptions and a meaningful relationship with supervisors, 56 respondents with limited work experience were excluded. This resulted in a final sample of 260 participants, of which 125 (48.08%) were female and 134 (51.54%) were male. The average age of participants was 24 years (SD = 4.41). In terms of race, 177 participants (68.08%) were identified as Caucasian, 14 (5.38%) as African American, 21 (8.08%) as Hispanic, 38 (14.62%) as Asian, and 10 (3.85%) identified as ‘Other’.

The South Korean data were collected from 205 MBA students at a large public university in Korea. The participants were primarily working professionals with significant work experience. Of the 205 respondents, 37 (18%) were female and 168 (82%) were male. The average age of the participants was 32 years (SD = 2.28). In terms of race, 202 participants (98.5%) identified as Asian, 2 (1%) as Caucasian, and 1 (0.5%) selected ‘Other’.

The Korean survey was distributed following a careful translation process. The author initially translated the survey into Korean, which was then back-translated into English by a Korean Ph.D. student majoring in English in the United States. To ensure the accuracy of the translation, two graduate students who are native English speakers and two graduate students who are native Korean speakers met to discuss whether the meaning of the questions had been effectively conveyed in the Korean version of the survey. All questions were presented using a 5-point Likert scale.

Measures

Organizational justice. As recommended by Colquitt (Reference Colquitt2001), procedural justice was measured using the scales developed by Thibaut and Walker (Reference Thibaut and Walker1975) and Leventhal (Reference Leventhal, Gergen, Greenberg and Willis1980). Specifically, I included 7 questions to measure procedural justice: 2 items from Thibaut and Walker (Reference Thibaut and Walker1975) and 5 items from Leventhal (Reference Leventhal, Gergen, Greenberg and Willis1980). Distributive justice was measured using the scale developed by Leventhal (Reference Leventhal, Berkowitz and Walster1976), which consists of 4 items to assess this dimension. Cronbach’s alphas were 0.83for procedural justice and 0.76 for distributive justice.

Leader–member exchange (LMX). The LMX-MDM scale (Liden & Maslyn, Reference Liden and Maslyn1998), consisting of 12 items, was used to measure LMX. Sample items include, ‘My supervisor is the kind of person one would like to have as a friend’ and ‘I do work for my supervisor that goes beyond what is specified in my job description’. Cronbach’s alpha was 0.92.

Work locus of control (WLOC). The scale for measuring an individual’s locus of control was originally developed by Rotter (Reference Rotter1966) and it has traditionally been the dominant instrument for measuring LOC. However, Phares (Reference Phares1976) criticized the Rotter Scale for its lack of precision and argued for the development of domain-specific measures of LOC. In response to this, Spector (Reference Spector1988) introduced the LOC scale to be more relevant to individuals’ experiences in the organizational context. The WLOC scale measures both internal and external locus of control in the workplace. Cronbach’s alpha was 0.80.

Control variables. To minimize the disadvantage of including undergraduate students as subjects and to account for alternative explanations as well as a variety of different factors that may influence the outcomes due to different cultural background of samples from different countries in this study, I added 12 variables as control variables. Control variables are gender, age, experience in full time job, working hour, position duration, job duration in lifetime, longest duration of working in lifetime, and perceived power distance. Education, race, job type, and position type are controlled as dummy variables.

Analysis

I tested the proposed hypotheses using hierarchical regression analyses. In line with Baron and Kenny (Reference Baron and Kenny1986) approach to mediation, I first regressed the mediators and outcome variable on the predictor variable (WLOC), followed by models that included the proposed mediators. Control variables were entered in the initial step of each regression to isolate the unique effects of the predictor and mediators.

This approach allows for the inclusion of multiple control variables and is well-suited for our study given the adequate reliability of our measures. As Jaccard and Wan (Reference Jaccard and Wan1996) noted, multiple regression is preferable to structural equation modeling (SEM) when measurement error is not a significant concern. Given our study’s focus on testing mediation effects with several control variables and acceptable reliabilities, Baron and Kenny’s framework was deemed appropriate.

Before conducting the mediation analyses, I assessed multicollinearity among the predictors and mediators to ensure that inflated standard errors would not compromise the interpretation of regression coefficients. The minimum, maximum, and mean VIF values are 1.25, 2.13, and 1.87, respectively.