The Mediating Effects of Organizational Commitment on Leadership Styles and Employee Performance in SOEs in Ghana: A Structural Equation Modeling Analysis

The primary purpose of this study is to investigate the mediating effects of organizational commitment on leadership styles (transformational, transactional and laissez-faire) and subordinate performance in state-owned enterprises in Ghana. The study adopted the partial least squares structural equation model (PLS-SEM) to test the hypotheses model. The model has been tested with 330 respondents. The data have been validated by the use of measurement modeling to determine internal consistency reliability (CR) and convergent validity (CV) with SmartPLS 3.0. The results indicate that organizational commitment mediates transformational and laissez-faire leadership. It has also been found that transformational and laissez-faire leadership have a statistically positive and significant relationship with job performance at p < .05 while transactional leadership proved insignificant. The implication is that leaders should know any action to improve subordinate commitment and job performance should take into account appropriate leadership behavior. Also, human development training should be instituted to shape the present and future leadership needs of the enterprise. In furtherance to this, the study provides discussions on the findings, limitations and went on to propose directions for future research.


Introduction
In this era of competitive business environment where profitability, efficiency, and effectiveness rules, employees should be influenced to increase their performance levels to either be in the organization or leave. For the public sector organizations (PSOs) to survive in this turbulent environment and achieve set objectives, they need to provide leadership that will influence subordinate job performance. Human capital is one of the most valuable resources of any enterprise, which does not only improve outcome but also offer a competitive advantage. Therefore, to improve the performance of employee is one of the essential objectives of contemporary public administration (PA) research suggestion (Jensen et al., 2016;Van Wart, 2013). Kehoe and Wright (2013) reveal that researchers and practitioners have realized the essence of employee performance through effective leadership in an organization. The leadership of the organization supposed to do everything necessary to inspire subordinates to improve their performance to achieve organizational success. The competence, behavior, trait, style, and approach of the leader are the major components that affect employee performance (Thigpen & Beauchair, 2012). An et al. (2019) also postulate that organizational performance is related to leadership. These and many other reasons suggest that leadership role in employee job performance is so pivotal that the business world cannot take its eye of it (Ruiz et al., 2011). Martina et al. (2012) reveal how important it is for leaders to take a keen interest in the needs and concerns of their talented human capital. In the same vein, the success of every organization, either a private, or public, is based on skill, knowledge, competence, loyalty, and contribution of subordinates toward work performance, which in turn produces the prodigious individuals which further enhances economic growth (Naz et al., 2012). This leadership styles interestingly affects the performance of the employee (Bacha, 2014;Donkor & Zhou, 2020;Masa'deh et al., 2016;Tse & Chiu, 2014) because the contribution of these employees is managed and controlled by the leaders in the organization. These and many more research works point to the fact that employee job performance depends mainly on the effectual leadership style.
The question of how leadership may affect employee performance is still gasping for a satisfactory response from researchers because there have been fragmentations in dealing with the full range leadership as instituted by Bass and Avolio (2004). For instance, research findings and conclusions on transformational and transactional leadership styles vary from one researcher and geographic to another Jia et al., 2017;Mclaggan & Botha, 2013). Again, some of the studies reiterated the importance of advancing on various methods to understand the exact relationship between employee job performance, commitment, and job satisfaction with other variables such as transformational, organizational citizenship behavior, culture and in few situations full range leadership styles (Afsar et al., 2016;Jia et al., 2017;Mclaggan & Botha, 2013;T. Park & Pierce, 2020;E. Park & Seo, 2016). Extant literature has shown that there are inconsistencies in their findings and thus making further research necessary in this direction in the public sector.
Our argument is based on the fact that there is a psychological mechanism that propels employee to perform efficiently and effectively through the exhibition of leadership styles. This mechanism needs to be understood by the leaders so that it may serve as a motivating factor to promote OC among employee (Chen et al., 2019) and to understand how leadership functions (Dust et al., 2014;Liao et al., 2016). This psychological mechanism creates work relationships between the employee and the leaders. The interpersonal relationship motivates employees to develop a positive workplace relationship (Chernyak-Hai & Rabenu, 2018). The critical paradigm here is social exchange theory (SET), whose basic premise is that human relations are formed based on subjective-cost benefit analysis which holds that people tend to repeat actions rewarded in the past and creates reciprocal gain-sharing environment with employees. This improves employee performance (Chernyak-Hai & Rabenu, 2018;Donkor & Zhou, 2020;Ugaddan & Park, 2017;Xerri, 2013). Again, this psychological mechanism arouses employee's passion, inclusion, and enthusiasm to help achieve organizational objectives (Huang et al., 2018). This further encourages employees to exceed job requirements and exhibits extra role behavior and pre-empt personal initiatives to achieve objectives and further improves employee performance (Karabiyik & Korumaz, 2014;Xerri, 2013). Therefore, employees recognize the organization as their own creation and have positive affection and commitment for its to be successful (Segal & Lehrer, 2012). When employees are committed to an organization, they become more attached and willing to continue to be part of the organization for a longer period (Uddin et al., 2019). Therefore, from SET perspective, organizational commitment will have a mediating and positive effect on employee performance.
However, several studies have examined the mediating variables between transformational and transactional leadership styles and other outcomes such as psychological empowerment, organizational citizenship behavior (OCB), motivation, identification, self-efficacy, leader-member exchange, and organizational commitment (Afsar et al., 2016;Cho et al., 2019;Jyoti & Bhau, 2016;Tse & Chiu, 2014;Zhu & Akhtar, 2014), but only a few have examined the effects of laissez-faire leadership (LFL) construct together with transformational and transactional leadership constructs. For this reason, the researchers aim to find out (if any) the effects of the mediating variable, organizational commitment on the full-range leadership constructs as the independent variables, and employee job performance as the dependent variable. The extant literature has revealed that most of these researches have been conducted in private and multinational organizations and educational institutions in the developed economies (Garza et al., 2014;Jyoti & Bhau, 2016;Veliu et al., 2017;Wiza & Hanganipai, 2014). Theoretically, our research contributes to the extant literature in two folds. First, our findings contribute to the leadership literature by trying to bridge the research gap the transformational, transactional leadership on one hand and laissez-faire leadership on the other, which has created a paradox in the literature of leadership. Second, this research enhances subordinate performance literature that leadership constructs, coupled with organizational commitment, affect employee performance by bringing out the psychological mechanism that causes the relationship. Practically, our findings provide much information to public sector managers to understand the leadership strategy to adopt to inspire subordinates to go beyond their required responsibility. Again, the study will serve as a guideline for managers to vary their leadership styles to suit the situation so as to create industrial harmony while achieving set goals.

Transformational Leadership and Subordinate Job Performance
Transformational leadership has existed since the era of Downton in 1973, but its emergence was until Burns (1978) introduced a book entitled leadership (Barroso et al., 2008). Transformational leaders can influence subordinates through inspiration and self-interest development that in turn, transform them within the organization (Pieterse et al., 2009). This valuable addition to the subordinates ideas and interests by the leader motivates them to perform beyond expectation (Yukl, 2004). Bromley and Kirschner-Bromley (2007), describes it as a task undertaken to make subordinates motivated by engaging in much higher ideas and moral values. According to (Donkor & Zhou, 2018;Pieterse et al., 2009;Salman et al., 2011) leaders with these traits need to inspire subordinates to act in the interest of the enterprise but not for themselves. Bass and Riggio (2006) also posits that transformational leadership hovers around leaders' capacity to motivate and inspire subordinates to put the organizational interest above their interest. Caillier (2016) and Peng et al. (2019) further states that leaders with transformational leadership traits encourage subordinates to go after the status quo with a new perspective to achieve novelty in their job performance. They further assert that transformational leader strives on innovation as a pivot to stimulate employees to be inspired and allow individual development and growth (Caillier, 2014(Caillier, , 2016. According to Bass and Riggio (2006) as cited by Çetin et al. (2012) and Saleem (2015), there are four critical elements of transformational leadership style, namely idealized influence, inspirational motivation, intellectual stimulation, and individualized consideration. Ruggieri (2009) explained that individuals are influenced via trust, respect, reposing confidence in them as followers by allowing them to take a stand on difficult issues, showing conviction, being focused on objectives of the entity, committed to their jobs, have strong values to work and ethically being responsible. A situation where the enterprise can offer economic and noneconomic rewards to subordinates, the leader uses the reward to enhance the performance of the subordinates (Ahmad et al., 2014;Wang et al., 2005). Mittal and Dhar (2015) suggest that transformational leads to subordinate performance. Existing literature on empirical research has revealed that transformational leadership behaviors have a significant influence on employee performance (Grant, 2012;Ispas, 2012;Rubin et al., 2005). Wang et al. (2011) concluded in their research that transformational leadership predicts better contextual performance than others. Walumbwa and Hartnell (2011) supported that transformational leadership influences subordinate performance.

Transactional Leadership and Subordinate Job Performance
Transactional leadership, unlike the transformational leadership, creates a mutual relationship with subordinates, where the contribution of both sides is acknowledged (Mcmahon, 2010). Here, the leader distinguishes the difference between the subordinate and the leader. The leader sets the goals and makes sure the subordinates achieve them. The leader with transactional leadership traits sets the goals clearly and put down strategies to achieve them by the subordinates. Eliyana (2010) also posits that the emphasis on transactional leadership is the achievement of set goals. This makes transactional leadership more practical. This leadership style is said to be most effective if there are clearly defined goals without any alternative to accomplish the task. Caillier (2014) states that unlike transformational leadership, transactional leadership does not create an opportunity for innovation, creativity but the risk-taking propensity is the primary concern for the subordinates to complete the task assigned them. Hence, the activities of the subordinates are task oriented. Therefore, novelty and new ideas are not encouraged but rather stick to the status quo. Burns (1978) further asserts that the primary concern of leaders with transactional leadership traits is not about how quality the work is but rather the quantity of the job performed and how to avoid resistance and make decisions effectively. This implies that transactional leaders do not show much concern about individuals nor their developmental needs but rather what they can offer to achieve organizational objectives is their prime concern (Northouse, 2015;Wenzel, 2007). To be effective, a transactional leader should recognize and monitor the accomplishment of employees and offer the necessary rewards to motivate them. Through this, monitoring mistakes and noncompliance could be identified, and correctional measures are applied. Pieterse et al. (2009) further state that transactional leader has the characteristics of remaining with the status quo to avoid risk. Transactional leadership is useful for improving leader effectiveness which engenders positive attitudes and performance among subordinates with contingent reward (Dai et al., 2013;Vito et al., 2014). Again, most research findings also support the fact that transactional leadership style has a positive relationship with subordinate job performance (Raja & Palanichamy, 2011). Wang et al. (2011) found transactional leadership to have better predictive individual job performance.

Laissez-Faire Leadership Style and Subordinate Job Performance
This leadership style can also be termed as "hands-off" or "free reign" style of leadership. The leader minimizes his involvement in decision making (Dubrin, 2007). Employees are given free rein to make decisions, and they are at liberty to change the processes and strategies. The laissez-faire leadership style is useful when subordinates are motivated to make effective decisions (Rowe & Guerrero, 2011). Donkor and Zhou (2020) added that in a situation where the employees are highly skilled, the laissez-faire leadership approach is practical. However, Yukl (2006) thinks that it is the absence of an effective leader. In this regard, there is no intervention, no follow-ups, disregard problems, and avoid making decisions (Gill, 2011). Two factors cause laissez-faire leadership. First, when there is a strong believe on the part of the leader that the employees know their job and for that matter leave them alone to do their job. Second, where the leader by fears that he may not be re-elected if he leads effectively if the position is elected, he may not desire to exert power and control. In such a situation, the leader provides minimal information and resources and fails to participate or involve in the workforce (Amoako-Asiedu & Obuobisa-Darko, 2017; Chaudhry & Javed, 2012). Research by Ghorbanian et al. (2012) revealed that laissez-faire leadership plays a significant role in many professions such as emergency medical services; therefore, in such instances, research supports the laissez-faire leadership. In Somalia, Ali and Dahie (2015) concluded that laissez-faire leadership has an impact on subordinate job satisfaction in public educational institutions. Babalola (2016) also found laissez-faire leadership as having a significant influence on media practitioners in Nigeria. However, Koech and Namusonge (2012) found that laissezfaire leadership not to have any positive relationship with subordinate performance. This suggests that state-owned enterprises in Kenya are not familiar with the laissez-faire leadership style.

The Mediating Role of Organizational Commitment
Many researchers resorted to organizational commitment as a mediator to examine the effects of transformational and transactional leadership on different employee behaviors including but not limited to performance (E. Park & Seo, 2016;H. Yeh & Hong, 2012), job satisfaction and turnover intention (Gyensare et al., 2017) motivation (Aslam & Khan, 2011;Binfor et al., 2013;Zareen & Razzaq, 2013). However, there is a dearth of studies on the analysis of the mediating effect of organizational commitment on the relationship among the full range of leadership constructs and work outcome concerning subordinate job performance. Again, extant literature suggests that effective leadership depends typically on individual circumstances; given this, we think that direct effects of the full range leadership constructs on employee performance may not give us a better comprehension unless we examine the role of mediating variable. Organizational commitment is a valuable employee variable because a highly committed subordinate at all times identifies with the objectives and values of the enterprise, craving to be with the entity and is amenable to go over and beyond their required job tasks (Ramshida & Manikandan, 2013). Therefore, a committed employee is an organizational competitive advantage. Barron and Chou (2016) suggest that employee commitment is the proper indicator of work performance and that leaders and supervisors should pay keen attention to their commitment through reciprocity tendencies. Choi et al. (2015) further explained organizational commitment as discernment and beliefs in which a subordinate accepts the values, mission, vision, objection, and goals of the organization and amenable to work and continue to be recognized as part and parcel of the enterprise. Again, Peng et al. (2019) suggest that organizational commitment is a force that will secure a follower to a course of action, which is of importance to one or more targets. Given this, organizational commitment is attributed to critical work-related factors such as absenteeism, labor turnover, performance, and loyalty to the organization (Alkhatani, 2016;Donkor & Zhou, 2020;Peng et al., 2019). Meyer and Allen (1991) again proposed a three-component model, defined in terms of desire, that is the employee wants to remain in the organization and support achieve objectives, need; that the employee needs to remain in the organization and support achieve objectives for identified reasons and obligation; that the employee must stay with the organization. These three components are described psychologically as affective, normative and continuance commitment which has been predominantly researched on in the subject of organizational commitment (Afshari & Gibson, 2016;Ennis et al., 2018;Gelaidan & Ahmad, 2013;Leroy et al., 2012;Meyer et al., 2002;T. Park & Pierce, 2020). Jaros (2012) believe that employee may be experiencing some psychic cost like a guilty conscience that would be incurred for breaking this obligation and leaves the organization. Entity managers and leaders need specific measures that may increase organizational commitment within the entity (Al Ariss et al., 2014). Farjad and Varnous (2013) support the fact that health, security, work conditions, and human development capabilities are among the factors that affect organizational commitment positively. Dey (2012) suggests the confident practice of the employer procedural justice within the organization (Gumusluoglu et al., 2013). Moreover, job satisfaction will positively increase organizational commitment (Gallato et al., 2012;Qamar, 2012).

Effects of Organizational Commitment on Job Performance
There have been several pieces of literature that have looked into the organizational commitment constructs to determine their relationship with employee performance in the organization. For an organization to achieve its objective, the subordinates must be much more committed to working toward that. Amoako-Asiedu and Obuobisa-Darko (2017) concluded in their studies that there is a relationship between organizational commitment and subordinates job performance. Various writers and scholars have resorted to the use of different procedures to determine that affective commitment has a positive relationship with job performance (Amunkete & Rothmann, 2015;Ferreira et al., 2013;Saha, 2016; see also Wu et al., 2011;H. Yeh & Hong, 2012). Other studies have also shown that the use of multidimensional methods has proven no relationship between continuance commitment and job performance (Meyer et al., 2002). Many other studies have also concluded that organizational commitment has a positive relationship with subordinate job performance (Amoako-Asiedu & Obuobisa-Darko, 2017; Çetin et al., 2012;Donkor & Zhou, 2020;Jackson et al., 2013;Long et al., 2014;E. Park & Seo, 2016;Saleem, 2015).

Leadership Styles, Organizational Commitment, and Subordinate Performance
Many scholars have found in many types of research that the performance of employees in an organization is somehow affected by a mediator or mediators. In this study, the researchers want to find how organizational commitment as a mediating variable affects employee performance. In our search, we realized among other things that many findings have stated that organizational commitment affects organizational activities and how it performs; C. Lee & Chen, 2013;Qaisar et al., 2012). Caillier and Sa (2016) found both transformational and transactional leadership to have a positive and significant relationship with commitment. Chi et al. (2012) found that leadership styles and organizational commitment have a positive and significant relationship with job performance. In Somalia, concluded that laissez-faire leadership has an impact on subordinate job satisfaction in public educational institutions. Babalola (2016) also found laissez-faire leadership as having a significant influence on media practitioners in Nigeria. In South Africa, Wiza and Hanganipai (2014) found among the academics similar results between transformational leadership style and affective and continuance commitment as having significant positive relations but transactional leadership had significant relations with normative commitment. In Kenya, the study concluded that there should be further studies to conclude the relationship between leadership styles and organizational commitment.
Concerning the public sector, Liou and Nyhan (1994) revealed that effective public sector leaders should pay attention to affective commitment to inspire the subordinates to work beyond expectation. Goulet and Frank (2002) studied among employees in public, non-profit, and for-profit sectors and revealed that public sector among the three groups was least committed to their various organizations. Again, Makhathini and Dyk (2018) agree with the low commitment in the public sector employees. The study found, among other things, that where there is role conflict, and ambiguity in work responsibilities, the level of organizational commitment is minimal. In sharp contrast, Ennis et al. (2018) found that public sector employees have a higher affective commitment and intrinsic satisfaction than their counterparts in the private sector. However, normative commitment and intrinsic satisfaction proved otherwise, but in aggregating the variables, one can deduce much stronger public sector employee relationship with organizational commitment and job satisfaction than the private sector. For effective performance by the employees, leaders must lay down enough factors to attract employees to work to attain the set objectives (Dey, 2012;Donkor & Zhou, 2020;Farjad & Varnous, 2013;Gallato et al., 2012;Gumusluoglu et al., 2013;Qamar, 2012). Yiing et al. (2009) concluded that organizational commitment as a mediator would improve subordinate job performance. Therefore, leadership influences employee commitment. So, when leaders portray a flexible leadership style, it will increase organizational commitment (Y. Yeh, 2014). In all the studies analyzed so far, none of them have scientifically or propounded any theory that examines the exact relationship between leadership styles and subordinate performance. Given the arguments and discussions adduced above, we hypothesize that:

Participants and Procedure
The data for this research were gathered from subordinates of 10 different organizations whose jurisdiction fall under the state-owned enterprises (SOEs) by the State Enterprise Commission of Ghana using paper questionnaires. This was done to get a more profound and proper understanding of the general relationship structure of the constructs. All the respondents were subordinates in various hierarchies in various departments of the organizations. The administrators of the questionnaires ensured anonymity and confidentiality of the targeted group as victimization is common in public sector organizations in that part of the world. The employees were assessed based on their employment-full-time employees in the state-owned enterprises in Kumasi Metropolis and Sunyani Municipality of Ghana. A total of 400 questionnaires were sent out to respondents, and 30 was unable to be returned with another 40 unusable because it lacked specific information that was very relevant for analysis. The descriptive statistics indicated that out of 330 respondents, 58.48% (193 respondents) are males. Most of the respondents (46.97%) are between the ages of 30 and 39 years of age. Majority of these public sector employees have experienced between 6 and 10 years. About 57% of these employees are a bachelor's degree holders and 14.85% are a master's degree certificate holders, and 46.97% of them are married. Again, the majority of these respondents being 55.15% have been under a supervisor for up to 3 years. Comparing it with the working length of these employees, one can conclude that internal transfers are frequent (Table 1).
Research tool and measurement procedure. The research instrument used for collecting primary data was Multifactor Leadership Questionnaire (MLQ) by Bass and Avolio (2004) and Meyer and Allen (2004). Three Components model Employee Commitment Survey Questionnaire (OCQ) which is seen to be most widely used data collection methods in evaluation research as well as personal performance questionnaire adopted from Yousef (2000). The Questionnaires used for the final data collection were close-ended questionnaires. Questionnaires help gather information on attitudes, opinions, behaviors, facts, and other information.

Multifactor Leadership Questionnaire (MLQ-5X Short)
To achieve the objectives validly, the researcher used a Multifactor Leadership Questionnaire introduced by Bass and Avolio to evaluate leadership styles and subordinate performance as one of the components of the questionnaire. This revised MLQ has five (5)-point Likert-type scale with 45 items that evaluate leaders and subordinates job performance. It consists of two rater forms of which the one that subordinates use to assess leaders was considered. The other is for self-evaluation by the leader (Bass & Avolio, 2004). MLQ has been extensively used to measure the full range of leadership styles and performance of leadership. Transformational leadership has four (4) dimensions, which include Idealized Influence, Inspirational Motivation, Intellectual Stimulation, and Individualized Consideration. Transactional leadership also has three (3) critical elements, including Contingent Reward (CR), Management By Exception (Active)(MBEA) and Management By Exception (Passive)(MBEP). No dimension has yet been attributable to laissez-faire leadership.

Organizational Commitment Questionnaire (OCQ)
The mediating variable was evaluated by the use of questionnaires introduced by Meyer and Allen (2004) known as the Three-Component Model (TCM). Employee Commitment Survey (revised version) has been extensively used to measure three forms of employee commitment of an organization. The survey included three well-validated scales, the Affective Commitment Scale (ACS), the Normative Commitment Scale (NCS), and the Continuance Commitment Scale (CCS). Each was scored separately to identify the commitment level of each employee.

Personal Performance Questionnaire
This questionnaire was adapted from Yousef (2000) and are in two dimensions, which included quality of performance and productivity measure. There are four items which are self-evaluating in nature by the employee on his or her performance and productivity as compared with that of colleagues doing the same task. The first two items were used to evaluate the performance quality and how productive a subordinate is on his or her job and the other two questions solicited and evaluated subordinates work quality and how productive they are is compared with colleague employees performing a similar function. The researchers used a 5-point rater scale to rank the items from 1 = very low to 5 = very high. The items included "What is the quality of your work?" "What is the productivity rate of your work? "Assess the work performance of your colleagues as compared to yourself?" "Assess your work performance in comparison with your colleagues who are doing the same work" (Yousef, 2000).

Respondents Demographic Questionnaire (RDQ)
The respondents' demographic information in the form of questionnaires were also collected. These included age, gender, promotional intervals, length of time working under the current supervisor, and educational qualification. If they have several current occupations (positions within the department), these respondents' questionnaires were structured and avoided any identity of the respondents to ensure anonymity.

Test for Validity and Reliability
This study adopted a scale which has been used and validated by other researchers. As stated above, the researcher used MLQ-5X developed by Bass and Avolio (2004) to evaluate leadership style, OCQ developed by Meyer and Allen (2004), Personal Performance Questionnaires and Respondent Demographic Questionnaires to evaluate the independent, mediating and dependent variables. The researchers themselves checked the validity, clarity, and accuracy of the content of the questionnaire. All the analyses, independent t-test, a test of reliability, statistical techniques of common method variance (measurement model evaluation) and Structural Equation Modeling (SEM) were used to analyze the data by the use of SmartPLS 3.0. The scale used by Hair et al. (2017) was used to check the reliability of subordinates performance. Yousef (2000) asserts that reliability indicates the accuracy or precision of the measuring instrument. According to Field (2005) and Pallant (2016), a Cronbach's alpha higher than .7 is indicative that the instrument has internal consistency.

Partial Least Squares Structural Equation Modeling
(PLS-SEM) was adopted to test the hypotheses of the study using SmartPLS 3.0. A two-way approach including measurement modeling-Consistency Reliability (CR), Convergent Validity (CV), as well as a structural model (i.e., path coefficient, the coefficient of determination and effect size) was used (Anderson & Gerbing, 1988;Hair et al., 2017).

Measurement Model Evaluation Results
In assessing and evaluating this model, the researchers considered the internal consistency reliability, convergent validity, and discriminant validity. Internal consistency reliability was done via composite reliability (CR), which demonstrates whether the items used in the study are reliable (McNeish, However, values between 0.60 and 0.70 are also acceptable in an exploratory study (Bagozzi et al., 1991;Hair et al., 2014). Cronbach's alpha values have also estimated for reliability and validity. Table 1 shows that the Cronbach alpha values were within the cut-off point of .70 or higher (.709 to .852). Both composite reliability and Cronbach's alpha are being reported for upper and lower bounds to determine the actual reliability of the internal consistency reliability.
The evaluation results prove that the measurement model of the constructs yields acceptable internal consistency reliability and Convergent validity also correlates positively with alternative measures of the same construct (Hair et al., 2014). It can be assessed by taking into consideration the average variance extracted (AVE; Avkiran, 2018). Given the AVE associated with the outer loadings, an outer loading should be 0.708 or higher to achieve at least 50% variance (0.5). The results showed that most of the outer loadings were within the threshold value and the remaining just below the outer loadings were maintained as they satisfied the AVE score of 50% and above after deleting few of them. It must be emphasized here that 0.4 or higher is acceptable if it is exploratory research (Hulland, 1999). Again, the research design was a cross-sectional and it is likely to encounter multicollinearity in the findings. To avoid this, the variance inflation factor (VIF) was evaluated for all the variables in the research to determine the presence of multicollinearity. The variance inflation factor statistics is the ratio of the overall standardized variance over tolerance variance and it shows that a variable is redundant if VIF exceeds 10.0 (Kline, 2011). The available VIF in this research values are between 1.2 and 1.4, indicating that multicollinearity was not present.
We further reported discriminant validity (DV) based on Fornell-Larcker criterion and Heterotrait-Monotrait criterion. The Fornell-Larcker criterion suggests that the square root of the AVE of each construct is higher than the construct's highest correlation with any other constructs in the model. This is depicted in those highlighted (diagonal values) and the correlation between the constructs in the off-diagonal position (0.815, 0.772, 0.730, 0.729, and 0.733). Compared to Heterotrait-Monotrait (HTMT) ratio of correlation criterion, because it performs better than traditional approaches of discriminant validity assessment as the values are much lower than both Fornel-Larcker criterion and the cross-loadings using 0.85 as a suitable threshold level (Hair et al., 2017). This again is for the fact that even though these methods are used frequently in applied research, these methods do not allow reliably detecting discriminant validity issues.
Therefore, Discriminant Validity (DV) is better assessed using HTMT (Henseler et al., 2015). DV is used to measure how different other constructs are using an empirical standard (Hair et al., 2017). To achieve a satisfactory level of DV, Henseler et al. (2015) recommend HTMT value above 0.90. This means that the HTMT value above 0.90 depicts no DV, indicating that the constructs are conceptually similar. The outcome shows that all the constructs have HTMT score <0.90, and bootstrap confidence interval of 97.5% showed all corresponding values fallen below 1, thereby indicating that all constructs of the study (LAISSEZ, OC, PERFORM, TRANS, and TRANSACT) are conceptually different and empirically distinct from each other.

Assessment of Structural Modeling Path Coefficients
Having confirmed the measurement model, the structural model was then assessed to test the conceptual model consisting of seven (7) different theoretically driven hypotheses. Parameters such as coefficient of determination (R 2 ), path coefficient (hypotheses tests), and effect size (f 2 ) were estimated as suggested by Hair et al. (2017). R 2 assesses model's predictive power (Hair et al., 2010). R 2 = 0.75, 0.50, and 0.25 are, respectively, described as "substantial, moderate, and weak" (Hair et al., 2014). The outcome of the results suggests that the R 2 is 0.169 for the performance endogenous latent variable. This means that the four latent variables (TRANS, TRANSACT, and LAISSEZ, and OC) weakly explain 16.9% of the variance in PERFORM. Also, TRANS, TRANSACT, and LAISSEZ together explain 19.1% of the variance of OC.
The inner model reveals that TRANS has the most potent effect on OC 0.441 followed by LAISSEZ (0.120) and TRANSACT (0.025). The hypothesized path relationship between LAISSEZ and OC shows that statistically, it is significant. Again, the hypothesized path relationship between TRANS and OC is statistically significant. However, about the endogenous variable, PERFORM, OC has the most potent effect on PERFORM (0.243) followed by LAISSEZ (0.220) and then TRANS (0.099) and TRANSACT (0.051). Again, the hypothesized path relationship between OC and PERFORM shows that statistically, it is significant, so do the hypothesized path relationship between LAISSEZ and PERFORM. Meanwhile, the hypothesized path relationship between TRANSACT and PERFORM is not statistically significant. This is because its standardized coefficient (0.060) is below the threshold of 0.1. Thus, we can conclude that LAISSEZ and TRANS are both moderately strong predictors of OC and again, LAISSEZ and OC are weakly predictors of PERFORM, but TRANSACT does not predict both the mediator and the endogenous dependent variable. f 2 , which measures specific exogenous construct's substantive impact on the endogenous construct (Hair et al., 2017) was then assessed after the R 2 . Cohen (1993) suggests that 0.02, 0.15, and 0.35 as respective guidelines indicating small, medium, and large effects were used to assess the effect size of the exogenous constructs. The results indicate that TRANS has a medium effect on OC (0.232), but the remaining constructs have small effects on OC. On PERFORM, even though LAISSEZ and OC have values above the small threshold, and they were below medium. TRANS and TRANSACT were also below the small level threshold. Therefore, the effect size on PERFORM was not significant. Stone-Geisser's Q 2 value has been recommended by Hair et al. (2014) for measuring the model's predictive relevance. The Q 2 value is obtained using the blindfolding procedure with certain omission distance (OD) such that Q 2 > 0 indicates that the model has predictive relevance for a specific endogenous construct. In calculating Q 2 values in this research, OD of 9 was considered to obtain the Q 2 values of 0.136 and 0.123 for the endogenous variables OC and PERFORM.
The outcome in the table above reveals that our path coefficients are significant. The results indicate that all paths are statistically significant except TRANSACT. The structural path of the remaining constructs is significantly based on a two-tailed test at p < .05.
In Hypothesis 1, we predicted that organizational commitment mediates transformational leadership and subordinate job performance. As shown in the table above, it was statistically significant at (β = 0.053, t = 2.162, p < .031). This suggests that organizational commitment is fundamental in seeking overall employee performance. Again, Hypothesis 2 also predicted that there is a statistically positive and significant relationship between transformational leadership and employee job performance. This was also proved decisive at (β = 0.148, t = 2.558, p < .011) which support the effects proposed in Hypothesis 2; hence, it is validated.
Hypothesis 3 was also predicted that organizational commitment would not be statistically definite and significantly mediates transactional leadership and subordinate job performance. Table 4 above shows that the t-values and p-values were not within the threshold of 1.96 and <.05, respectively (β = 0.008, t = 0.748, p < .454); therefore, the hypothesis was accepted. The same could be seen with Hypothesis 4, which was again predicted that there is no statistically positive and significant relationship between transactional leadership and performance. The analysis revealed (β = 0.060, t = 0.956, p < .339). All the thresholds were missed, and for that matter, we accept the hypothesis.
Hypotheses 5 and 6 gave a fascinating revelation. While the indirect effect was rejected, the direct effect was accepted. That is, the mediated Hypothesis (5) was not statistically significant as the study did not meet all the thresholds. Therefore, Hypothesis 5 was rejected at β = 0.035, t = 1.893, p < .058. Even though not significant, the null hypothesis accepted. This suggests that the leaders in public sector organizations should vary their styles to motivate subordinates to be more committed to their job performance. Hypothesis 6, however, was accepted at β = 0.217, t = 3.747, p < .000. This revealed that mediating role is not significant in subordinate performance in an environment where the leader is practicing laissez-faire leadership style.
Hypothesis 7 was also predicted to have a statistically positive and significant relationship between organizational commitment and performance. Having fallen within the thresholds at β = 0.160, t = 2.415, p < .016, the hypothesis was accepted. This further suggests that organizational commitment is very fundamental in subordinate job performance. Therefore, leaders need to work on the style that attracts commitment from the subordinates.

Discussion
The primary purpose of this study was to examine the mediating effects of organizational commitment to leadership styles and subordinate job performance. The study was conducted in state-owned enterprises (SOEs) in Ghana, which to the perception of the researchers should have some peculiarity behavior and findings compared to studies in the developed nations. In all, the study is consistent with the existing research showing that state-owned enterprises and for that matter, public sector organizations are not different in terms of leadership approach no matter where the business is located. Extant literature on leadership and subordinates job performance have suggested that there is a positive relationship between transformational leadership, job performance, and organizational commitment (Amoako-Asiedu & Obuobisa-Darko, 2017;Donkor & Zhou, 2020;Long et al., 2014;Saleem, 2015;Wiza & Hanganipai, 2014). The findings of this study on transformational leadership have been consistent with all these research findings (Deluga & Souza, 2011;H. W. Lee, 2010;Makhathini & Dyk, 2018;Vogel & Masal, 2015). The results further revealed that transformational leadership sets the center stage for managerial circumstances to foster the growth of state-owned enterprises. This is again supported by extant literature that found transformational leadership to foster subordinate opportunity to be creative (Afsar et al., 2014(Afsar et al., , 2016Caillier, 2016).
Because of this, managers in state-owned enterprises may be most effective if they display transformational leadership trait by fully adopting the constructs of a transformational leadership style. This may cause subordinates to be much more loyal and committed to their chosen job, and in turn, improve organizational commitment. This shows that transformational leadership style is an essential vehicle in fostering subordinate commitment, which implies that subordinate is working harder without the intention to leave. Therefore, one can conclude that a transformational leadership style is a statistically positive predictor of organizational commitment and subordinates job performance.
Again, transactional leadership is said to be a negative predictor of employee commitment and job performance. Even though it is consistent with our hypothesis, other studies have found our findings otherwise. Alamir (2010) tests whether transformational and transactional leadership impact on job satisfaction. The correlational and regression analysis revealed that transactional leadership is positively related to organizational commitment. However, studies by Shrestha and Mishra (2011) in Nepal technology-based organization found an insignificant relationship between transactional leadership and organizational commitment. H. W. Lee (2010) further concluded to support the extant literature that transformational leadership and transactional leadership have a significant positive effect on organizational commitment. Also, Obasan and Banjo (2014) conclude that transactional leadership is negatively correlated with organizational commitment. Leaders who exhibit transactional leadership traits should find innovative ways to motivate employees to go beyond their job responsibilities to improve organizational commitment and performance. In Donkor and Zhou (2020), they concluded in their study that transactional leadership is detrimental to subordinate innovative work behavior. To maintain the status quo, these leaders should allow subordinate to operate freely and reduce interference to the lowest minimum to benefit the SOEs. The above findings suggest that even though transactional leadership in this study rejects the null hypothesis of the study when it is linked with transformational leadership, and the leaders motivate subordinates, they may be much more committed to achieving organizational goals. Also, the cause of this is attributed to the fact that the study was conducted in state-owned enterprises where most people do not give of their best, due to low motivation, therefore, do not accept this leadership style that seeks to put pressure and timelines on their job responsibilities. Again, laissez-faire leadership showed a different outcome in the direct and indirect effects. That is in terms of mediator variable; laissez-faire proved statistically insignificant while the direct effect on performance showed a statistically positive significant. Obasan and Banjo (2014) did a study that concluded that there is a negative correlation between laissez-faire leadership and organizational commitment to employee performance. They further state that applying it in an organization will negatively impact on the organizational behavior. Even though the relationship between laissez-faire and organizational commitment has not been examined extensively by many researchers and scholars (Jackson et al., 2013), there are some positive findings (Babalola, 2016;Cemaloglu et al., 2012;Ennis et al., 2018).
Organizational commitment has extensively discussed in the extant literature that supports our findings. Studies in both public and private sector organizations are found to have a statistically positive and significant effect on employee job performance (Jussila et al., 2012;Wu & Chen, 2018; see also, Donkor & Zhou, 2020;Meyer et al., 2002). Ennis et al. (2018) further concluded that leadership styles influence organizational commitment, and in turn, the organizational commitment will affect subordinate job performance and mediate the relationship between leadership style and employee job performance. These findings suggest that employees in public sector, like any other sector, have the will to continue to work in the organization, and so, the leadership have the responsibility to engender trust, motivation and inspire a sense of loyalty and duty. This will encourage subordinates to continue to give off their best for the achievement of objectives. Even though most organizations do not adhere to laissez-faire leadership for commitment, leaders should allow free rein, especially, where subordinates are professionals who can perform with little or no supervision to work on their way. This concludes that organizational success largely depends on how long the committed subordinate remains in the organization to reduce labor turnover and also understands the values and strategies of the enterprise. This creates strong ties and increases trustworthiness with clients and customers when they believe that employees at all times will be available to serve them. Hence, the inconsistencies in the extant literature have been found to have relationship with the geographic location and cultural backgrounds of the employees being studied.

Implications
Theoretically, our findings help to understand the main drivers of subordinate performance through the commitment of employees and suitable leadership styles exhibited by the leaders. It contributes to our understanding and psychology of subordinate attitude toward performance and effects of the full range of leadership styles in organizational behavior. As transformational leadership style fosters innovativeness among subordinates, transactional leadership also reward them with effective performance while laissez-faire leadership style allows for exhibition of professional competencies with little or no interference from the leader. Thus, the effectiveness of these leadership styles in encouraging subordinate performance is dependent on organizational commitment.
Practically, findings of the study imply that when subordinates perceive their leaders as adopting leadership behavior that allows them to take part in decision making, they become more committed to the enterprise, more loyal to the course of the entity and improve on their job performance. To this, leaders should bear in mind that any action to improve subordinate job performance and commitment should take into consideration the fact that subordinates' satisfaction and empowerment is essential. Again, in improving subordinate job performance and commitment, managers should adopt appropriate leadership behavior. Appropriate human development training programs should be developed to shape subordinate's leadership needs of the enterprises through the development of a succession plan that reduces the central government's interference. This will motivate subordinates to be more committed to the course of the state enterprises. After all, they may feel that the future is bright if they continue to improve on their performance. The study is no doubt has some implications for SOEs leaders who wish to stimulate and transform their entities to keep their valuable resources, human, for them to stay committed to reduce high labor turnover and also improve on productivity through working beyond responsibilities. It is essential for managers in public institutions to understand the leadership strategy necessary to inspire subordinates to go beyond their expectations to help achieve set goals. Subordinates who are committed to organizational objectives perform efficiently and effectively to achieve organizational objectives. Again, leaders should vary their leadership styles to suit the situation and further empower subordinates to take up higher responsibility through delegation and involvement in decision making especially on matters that bother on their organizational well-being and job security.

Limitations and Future Research
Despite the immense contributions of this research, it is bedeviled with a couple of setbacks that the researchers can work on in their future research endeavors. First, the study focuses on only state-owned enterprises in Ghana. Future research might consider the entire public sector organizations, which is much broader with more cost centers and departments. Second, the study was cross-sectional, which may not necessarily reflect the causal relationship among the variables. Future research should consider a longitudinal research design to examine the relationships among variables. Third, the data were collected from different organizations which fall within state enterprises commission in Ghana but failed to perform multiple level analysis. Finally, the researchers are encouraged to consider other mediated variables like job satisfaction, leader-member-exchanged (LMX) and organizational culture and further conduct longitudinal studies to explore the causal relationship among the variables.

Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding
The author(s) received no financial support for the research and/or authorship of this article.