THE EFFECT OF LEADER MEMBER EXCHANGE (LMX) AND TEAM MEMBER EXCHANGE (TMX) ON EMPLOYEE PERFORMANCE THROUGH AFFECTIVE COMMITMENT AT PT PERKEBUNAN NUSANTARA X

: This research aims to determine the effect of Leader Member Exchange (LMX) and Team Member Exchange (TMX) on employee performance through Affective Commitment. The population in this study were all employees of PT Perkebunan Nusantara X. The analytical tool used was Path Analysis by collecting basic data through questionnaires. From the results of the analysis, it can be concluded that LMX and TMX has a significant effect on positive affective commitment, while LMX has no significant positive effect on employee performance, TMX has a significant effect on employee performance, and affective commitment has a significant effect on employee performance. Therefore, affective commitment is a good intervening variable, especially on the effect of LMX and TMX toward employee performance. Our finding that the TMX differentiation mode feeds on the TMX employee performance relationship suggests that TMX differentiation should be included in a theoretical framework on employee performance. When employees have high TMX differentiation, different social exchange relationships among employee members jeopardize the role of TMX. These findings suggest that high TMX differentiation may impair the positive effects of TMX. In the company, the leader should pay attention to build a different relationship among the members of the employees. High TMX differentiation can lead to a lower sense of unity within employees, which in turn appears to be detrimental to the contribution of TMX work units to projected employee performance. The effects of the TMX differentiation concept reflect the leader's task tasks, employee work coordination,


Table 1. Financial Ratios
Source: Annual Report 2020 To overcome the decline in performance at PTPN X, the company wants its employees to have a high level of affective commitment; this is a momentum to build upon in the context of overall reform in an effort to accelerate business development, which results in improved service quality and the development of the company's mission as a State-Owned Enterprise, one of which has the task of being self-sufficient in food commodities, particularly sugar. In this situation, management is attempting to raise employee awareness through the phrase sense of crisis & sense of belonging, which is the fundamental value of production implementation in 2022. The Company's operational activities, in order to achieve the previously stated business objectives, require the presence of qualified Human Resources (HR) or Human Capital (HC). To produce excellent human resources, the Company's vision, mission, and values, as well as the work culture of PTPN X, serve as a roadmap for the Company's development. These guidelines serve as the basis for all PTPN X workers to ensure that they collaborate in the specified direction. The Company's work culture is AKHLAK; this work culture must be lived and applied by all PTPN X personnel to ensure that employee productivity and competitiveness stay high.
Furthermore, it is expected that this research will be able to clearly identify the factors influencing horizontal and vertical cooperation as measured by the influence of Leader Member Exchange (LMX) and Team Member Exchange (TMX) on employee performance through affective commitment, so that it can later be used as a reference by related parties, in this case the Head of Human Resources of PTPN X, to evaluate the level of LMX, TMX, and TMX. Additionally, it can be used as corrective material for employees' perceptions of working conditions and circumstances, including vertical and horizontal cooperation, decision-making procedures, reward systems, and the support provided by supervisors to their subordinates in order for employees to develop an emotional attachment to the company. Even when an employee can identify with the company, he or she will be able to contribute to the company's growth.
profession, however Graen & Uhl-Bien (1995) suggest that LMX consists of three components: leadership, subordinates, and interpersonal interactions. According to Taqiuddin et al. (2018), LMX has a positive and significant effect on employee performance because it enables leaders to recognize the potential of material that matches their capabilities, resulting in higher employee performance. Additionally, Shu & Lazatkhan (2017) claimed that LMX has an effect on employee performance, revealing that the relationship between superiors and subordinates is mutually correlated due to the presence of a high-quality relationship defined by trust, loyalty, and commitment to their work.
: There is a significant effect of LMX (X 1 ) on employee performance (Y) at PTPN X employees.

The Effect of TMX on Employee Performance
According to previous studies (Liden et al., 2000), TMX shows a positive and direct relationship on performance. Another study (Jordan et al., 2002) discovered that TMX had a positive effect on employee performance as measured by management. The importance of relationship quality is placed on assisting coworkers, sharing ideas, and providing feedback on work challenges encountered by coworkers (Liao et al., 2013). A high TMX quality reflects an individual's proclivity to cherish connections with coworkers, contribute ideas and abilities to the progress of work units, be cooperative, and show respect for coworkers (Banks et al., 2014). On the other hand, low TMX quality suggests that individuals in work units are less engaged with groups and less inclined to exchange knowledge, which might have an effect on group work processes (Liu, Keller, et al., 2011a).
: There is a significant effect of team exchange member (X 2 ) on employee performance (Y) at PTPN X employees.

The Effect of LMX on Affective Commitment
According to Morrow et al. (2005), LMX is an improvement in the quality of the relationship between supervision and employees will be able to improve the work of both. This can happen because the relationship between employees and superiors has a good communication relationship. Further, Lee (2005) have the same conclusion that increasing LMX can increase employee commitment to the organization.
: There is a significant effect of LMX (X 1 ) on affective commitment (Z) of PTPN X employees.

The Effect of TMX on Affective Commitment
Individual roles will be valuable to the group if each member has employees who are reciprocal and support each other, resulting in extra role behavior toward coworkers and other parties (Farmer et al. (2015); Seers (1989)). Helping behavior also implies that group members' interactions have been of high quality, demonstrating that group members have completed the assimilation process, had a sense of unity, and had trust among coworkers. Based on the foregoing, it can be stated that high-quality employee relationships can serve as a foundation and guide for individuals in action, as they can foster trust and strengthen relationships among members of the group. Furthermore, high-quality relationships among coworkers will produce a conducive work environment since coworkers' support and encourage one another, making them more likely to cooperate. Some previous studies have shown that the quality level of TMX is directly related to the affective commitment level of individuals in organizations (Chiaburu et al., 2013).
: There is a significant effect of TMX (X 2 ) on affective commitment (Z) of PTPN X employees.

The Effect of Affective Commitment and Employee Performance
Affective commitment refers to members' emotional attachment to the organization, identification with the organization, and participation in organizational activities. Organization members with high emotional commitment will remain members of the organization because they want to (Meyer, 1997). As a result, the study supports this viewpoint, especially (Liao et al., 2013) and (Lam, 2003) that high organizational commitment can have a positive impact on boosting employee performance.
: There is a significant effect of LMX (X 1 ) on employee performance (Y) through affective commitment (Z) to PTPN X employees. : There is a significant effect of LMX (X 2 ) on employee performance (Y) through affective commitment (Z) to PTPN X employees. : There is a significant effect of affective commitment (Z) on employee performance (Y) at PTPN X employees.

Research Method
This research applies a quantitative method. The types and sources of data used in this study are quantitative data in the form of answers provided by respondents to questionnaire questions. The research population consists of PTPN X employees from all work units (Head Office, Sugar Factory and Tobacco Gardens). Human Resources (HR) at PTPN X had a workforce of 8,208 individuals as of December 31, 2020 (permanent employees and temporary employees).
The sample size is 202 respondents. This study will employ the landing page method in the form of a questionnaire page on the internet login page, requiring each user who logs in to complete the questionnaire. T the data analysis technique used in this study was partial least squares (PLS). According to (Jaya, et al., 2008) the use of the PLS method is appropriate for this study because it does not require data with a specific distribution and is very flexible for this study because it does not require too many samples. However, because there are three mediating variables in this study that alter the relationship between the independent and dependent variables, mediation testing can be performed utilizing the Sobel test procedure. In this study, the independent variable is employee performance (Y), while the dependent variable is LMX (X 1 ) with TMX (X 2 ), and affective commitment (Z) as intervening variables.

Validity test
Validity test to assess the accuracy of the measuring instrument (questionnaire) (Sujarweni, 2018). In SEM analysis, the validity test looks at the results of the loading scores of indicators or statements from the questionnaire. There is a standard rule of Fonel et.al. (1992) in (Santosa, 2018), that the loading of each indicator to the corresponding latent variable is at least 0,7. Meanwhile, there are several other opinions regarding the value of loading, Chin (1998) in (Santosa, 2018) stated that a large loading value of > 0,6 and a small of < 0,7 can still be used  From the results of the data processing output with Smart PLS 3, the outer loading obtained with some values <0,5, so that they were deleted, namely X1.1, X1.4, X1.7 and X1.10. Furthermore, the data is reprocessed, the results are as follows in Table 3 and Figure 4. Outer loading results as a result of the validity test, show the loading factor value above > 0,70, but only one indicator is 0,679, this cannot be deleted because it is representative of indicator theory (Muhli (BPS), 2018), so it was concluded that all loading factors were declared valid.

Reliability Test
The reliability value of the CR construct is recommended > 0,7 (Sujarweni, 2018). To find out internal consistency by looking at the Cronbach's Alpha value, the minimum value of which is 0,7 (George and Mallery, 2003in (Santosa, 2018). The reliability test in PLS can use two methods, namely Cronbach's alpha and Composite reliability. Cronbach's alpha and Composite reliability values must be greater than > 0,7 although the value of 0,6 is still acceptable (Hair et.al, 2008in (Abdillah, 2015). From the results of the Smart PLS3 test obtained as follows: In Table below, it shows the value of Cronbach's Alpha and the value of composite reliability is above 0,7, so it can be concluded that all latent variables are declared reliable, so the questionnaire is feasible and can be used to measure latent variables. The results of the second model structure test are Leader Member Exchange (X1) on Employee Performance (Y) which obtained path coefficient of 0,127, it means that the direct

Hypothesis Testing Analysis
In Table below describes the hypothesis of a direct effect. Further, the results of the t-test of the hypothesis H1: Leader Member Exchange on Affective Commitment are obtained at 6,896, where the value is greater than the t table of 1,66. With the number of samples (n) = 100 and the number of independent variables (k) = 2. Significance using one side, with 5%, and Degree of freedom or degrees of freedom (df = nk), then df = 100 -2 = 98 which obtained t table 1,661, which indicates significant results. The results of the t-test of the hypothesis H2: Team Member Exchange on the Affective Commitment were obtained at 5,577 > 1,66, which was significant. The results of the t-test of the hypothesis H3: Leader Member Exchange on Employee Performance, the results of the t-test of 1,056 < 1,66 are not significant. H4 hypothesis t test results: Team Member Exchange on Employee Performance, the results of the t-test of 2,567> 1,66, which means significant results. Likewise, the results of the t-test of the hypothesis H5: Affective Commitment on Employee Performance, the value of the t-test results is 2,079> 1,66, which means the results are significantly positive. The results of the t-test of the hypothesis H6: Leader Member Exchange on Employee Performance through the Intervening Affective Commitment variable, the results of the t-test were 1,963 > 1,66, which means the significant path is positive. The results of the t-test of the hypothesis H7: Team Member Exchange on Employee Performance through the Intervening Affective Commitment variable, the results of the t-test of 1,838> 1,66, which means that the significant path is positive.

Discussion The Effect of Leader Member Exchange of PT Perkebunan Nusantara X on Employee Performance
From the results of the H1 hypothesis test, it was obtained that the Leader Member Exchange (X1) has no effect on Employee Performance (Y). The performance of the Leader Member Exchange of PT Perkebunan Nusantara X cannot guarantee PT Perkebunan Nusantara X's customers to be loyal, but this is related to whether the consumer is satisfied or dissatisfied with the performance of the Leader Member Exchange they receive. This result is in line with the research results by Kamila & Arwiyah (2019) that the Leader Member Exchange has no effect on employee performance, as well as research by Firdaus (2017); Ikhbar (2015), and Chen & Wei (2020). Besides that, this result is supported by Truckenbrodt (2000) which highlight that if the relationship between employees and leaders is well established, employees tend to be willing to work more than their share, work immediately, provide more initiative and work more optimally.
These results indicate that the increase or decrease in LMX will not have any effect on employee performance. Employees will continue to work according to their respective duties even though the perception of LMX is not good. This is in line with the research conducted by Arijadi (2017) who found that LMX is unable to improve employee performance. This research also corroborates this strengthens the theory by Landy (1989) who argues that the interaction between superiors and subordinates is of low quality, it will be characterized by: (1) The interaction between superiors and subordinates tends to be formal, therefore a superior will use the power of authority in influencing his or her subordinates; (2) The existence of a low level of role negotiation so that the interaction between superiors and subordinates is limited by the roles played; (3) The superior rarely talks with the subordinate about the effectiveness of the task; and (4) Superiors rarely help subordinates by giving different tasks.

The Effect of PT Perkebunan Nusantara X's Team Member Exchange on Employee Performance
From the results of hypothesis testing H2 that obtained Team Member Exchange (X2) has an effect on Employee Performance (Y). This means that if Team Member Exchange increases, it will directly increase Employee Performance as well. Trust is a strong consumer conviction in the company. If consumers trust the company, PT Perkebunan Nusantara X's consumers will be strong and loyal, so they will not turn to other companies. This is what makes hypothesis H2 accepted. The results of the study are supported by previous research (Liu, Keller, et al., 2011b).
Our finding that the TMX differentiation mode feeds on the TMX employee performance relationship suggests that TMX differentiation should be included in a theoretical framework on employee performance. When employees have high TMX differentiation, different social exchange relationships among employee members jeopardize the role of TMX. These findings suggest that high TMX differentiation may impair the positive effects of TMX. In the company, the leader should pay attention to build a different relationship among the members of the employees. High TMX differentiation can lead to a lower sense of unity within employees, which in turn appears to be detrimental to the contribution of TMX work units to projected employee performance. The effects of the TMX differentiation concept reflect the leader's task tasks, employee work coordination,

The Effect of Leader Member Exchange of PT Perkebunan Nusantara X on Affective Commitment
From the results of the t-test of the H3 hypothesis, it was obtained that the Leader Member Exchange (X1) directly had a significant positive effect on the Affective Commitment (Z) variable. This gives an indication that if the Leader Member Exchange is good, it will directly increase the Affective Commitment. PT Perkebunan Nusantara X, in this case has succeeded in providing services that can provide PT Perkebunan Nusantara X's consumer expectations as great as what consumers feel after placing an order, so that satisfaction is created, this supports the hypothesis H1 is accepted, namely the Leader Member Exchange of PT Perkebunan Nusantara X has an effect on Affective Commitment. This study provides the same results as previous research (Hardiyana et al., 2021;Harthantyo & Rahardjo, 2017;Ikhbar, 2015). Therefore, the better the implementation of leader-member exchange at PT Perkebunan Nusantara X, the better the affective commitment of the employees as well. Other studies have found that employee engagement with the organization will increase when there is a good quality relationship between members in the organization and their leaders (Son, 2015).

The Effect of Team Member Exchange PT Perkebunan Nusantara X on Affective Commitment
From the results of hypothesis testing H4 it was found that Team Member Exchange (X2) had a significant positive effect on Affective Commitment (Z). This shows that if Team Member Exchange increases, it will directly increase Affective Commitment as well. Creating and maintaining Team Member Exchange is through maintaining good relationships with customers in the long term. For this reason, the company is trying to make PT Perkebunan Nusantara X consumers believe in the company, thus creating an Affective Commitment, this supports the accepted hypothesis H2, namely Team Member Exchange PT Perkebunan Nusantara X has an effect on Affective Commitment. This is in accordance with research by Kim et al. (2018) and Liu, Loi, et al. (2011) increases, it will directly increase Employee Performance as well. If the desire is in accordance with reality, it means that the Affective Commitment from PT Perkebunan Nusantara X is achieved, it will have a direct impact on increasing employee performance. Research results are supported by Arijadi (2017).
Other research shows that organizational commitment has a significant effect on employee performance (Folorunso et al., 2014), which means that the higher the organizational commitment, the higher the employee's performance. This finding supports the opinion of (Pramandhika & Mas'ud, 2011) that states employee loyalty or commitment to the company is seen as very important in business. Employees who have high loyalty or commitment will perform well and prioritize the interests of the company over their own interests.

Leader Member Exchange PT Perkebunan Nusantara X on Employee Performance through Affective Commitment
In Table 7, H6 shows the indirect effect of PT Perkebunan Nusantara X's Leader Member Exchange on Employee Performance is greater than the direct effect, which means that the Affective Commitment variable can act as an intervening variable between Leader Member Exchange and Employee Performance. It is very important to fulfill the Leader Member Exchange in accordance with expectations and reality, so that Affective Commitment is achieved which will have implications for Employee Performance. This is related to the H1 hypothesis where the Leader Member Exchange cannot partially affect employee performance, but with the mediating variable Affective Commitment, the Leader Member Exchange can have a more positive impact on employee performance. This research is supported by the results of Lie Darwin's research et.al that Leader Member Exchange influences Employee Performance through Affective Commitment. Further, Arijadi (2017) highlight that the desire to stay in the company makes employee performance increase. Organizational commitment is able to mediate between the influence of member exchange leaders (LMX) on employee performance (Herlambang, 2017).

Team Member Exchange PT Perkebunan Nusantara X on Employee Performance through Affective Commitment
In Table 7, H7 shows that the direct effect of PT Perkebunan Nusantara X's Team Member Exchange on Employee Performance is greater (0,355 > 0,107) than indirectly through Affective Commitment. It means that Affective Commitment cannot act as a mediation between Team Member Exchange and Employee Performance. This shows that in order to further improve Employee Performance, it is better to be able to directly from Team Member Exchange for Customer Loyalty without being mediated by Affective Commitment.

Conclusion
To sum up everything that has been stated so far, the following conclusion can be drawn in this study, namely: LMX and TMX have a significant positive effect on affective commitment, while LMX has no significant positive effect on employee performance, TMX has a significant positive effect on employee performance and affective commitment has a significant positive effect on employee performance. According to this result, it can be said that affective commitment is a good intervening variable, especially on the effect of LMX and TMX on employee performance.