The Effect of Leadership and Organizational Justice in Predicting Positive Emotion of Indonesian Civil Servants dalam Aparatur Sipil Negara Indonesia

This study aims to explore the influence of leadership styles (transactional and transformational) on organizational justice (distributive justice and procedural justice) and their impact on the positive emotions of Government civil servants,which is located in the City of South Tangerang, Palu, and Sorong. In the government environment, perceptions of leadership styles (transactional and transformational) and organizational justice (distributive and procedural) possessed by employees can vary. Thus, itindeed tends to influence the formation of positive emotions. The Social Desirability Response (SDR) test was examined in the validation test. A survey method with a purposive sampling technique (n = 400) was used to collect the data. The results showed that both transformational and transactional leadership styles have a positive and significant effect on procedural justice and distributive justice. Hence, having a subsequent impact on the formation of positive emotions


INTRODUCTION
Human resources are one of the critical assets in the organization, both business and government organizations. This because human resources play a role as subjects of implementing policies and operational activities of the organization or company. Resources owned by organizations or companies such as capital, methods, and machines will not provide optimal results, if not supported by human resources with their optimal performance (Swanson & Holton, 2001).
Effective leadership is an integral part of organizational effectiveness, especially in government organizations (Ingraham & Getha-Taylor, 2004). In government organizations, effective leaders will create a positive organizational culture, strengthen motivation, clarify the mission and goals of the organization, and direct the organization to more productive and highperformance outcomes (Ingraham & Getha-Taylor, 2004).
In a government organization, the success or failure of carrying out its duties and operations is greatly influenced by leadership factors. The implementation of good governance will be realized if supported by the adequate capacity of government organizations. However, Thoha (2008) stated that weakness in the leadership factor is one of the main factors causing the fall of bureaucratic performance in Indonesia.
The statement is consistent with Crosby and Bryson (2018) that in various studies on leadership in government organizations, always focused on transformational leadership and transactional leadership. In government organizations, there are always relationships that involve exchanges between superiors and subordinates so that affection and motivation are often created by fellow employees.
Besides, an organization often found differences between employees with other employees, both in terms of the treatment of leaders to subordinates, the provision of wages, or policies issued by the leadership. It will cause the perception of organizational justice related to the feelings of an employee regarding the distribution of rewards or incentives, the allocation process, and the way they are treated in an organization (Colquitt et al., 2001).
In government organizations, there are always organizational injustices that are felt by subordinates as a result of leadership treatment, resulting in positive and negative emotions (Budiyanti et al., 2018). The emergence of this phenomenon, in general, is caused by several factors that have become common knowledge in government organizations, such for example the distribution of official travel and the distribution of incentive activities (Budiyanti et al., 2018). Things like these often lead to the formation of positive and negative emotions of civil servants that have an impact on job satisfaction.
Several studies have shown that organizational leadership and justice have an essential role in an organization (Wang et al., 2010;Guh et al., 2013;Chen & Jin, 2014;Kim et al., 2017;Unterhitzenberger & Bryde, 2019). Furthermore, Iqbal et al. (2018) also stated that the importance of organizational justice in implementing leadership styles in an organization. Managers who do not understand the importance of organizational justice would be deemed not to have theright leadership style and would be boycotted by their followers (Sharma, 2016). It is further confirmed that organizational citizenship behavior in the context of social exchange with the quality of the relationship between leaders and subordinates shows a close relationship between leadership and organizational justice (Cropanzano et al., 2002;Armagan & Erzen, 2015;Chen et al., 2018).
While according to Colquitt et al. (2001), severalstudies in the field of organizational justice shows that the perception of justice has an essential influence on organizational commitment, job satisfaction, intention to quit work, organizational citizenship behavior, and performance. It is acknowledged that considerations of justice can cause emotional responses, especially in times of injustice (Cohen-Charash & Spector, 2001;Barclay et al., 2005;Cropanzano et al., 2007;Wu et al., 2017).
Previous justice theories and other studies showing that emotions are part of the relationship between experience due to injustice and the tendency to retaliate (Barclay et al., 2005;Tracy et al., 2007;Skarlicki et al., 2008), however only a few studies have tested regarding the relationship between justice and emotions. The results of the literature review conducted by the author show that organizational justice consisting of procedural justice, distributive justice, and interactional justice has an influence on emotions and has an impact on job satisfaction. These studies include previous research (Cohen-Charash & Spector, 2001;Rupp et al., 2006;Mazurkiewicz, 2009;Gotlib, 2011).
In general, in government organizations, civil servants will compare the results obtained with their fellow employees, so that policy implementation is perceived as having justice among them. In most organizations, such as government organizations, always have the view that when employees receive satisfactory results, organizational justice has been created (Cropanzano et al., 2008). It is because most superiors cannot distinguish the desired results from the expected results (Cropanzano et al., 2008;Cropanzano & Stein, 2009). In other words, the results desired by most employees are not necessarily fair. It is consistent with research conducted by (Cropanzano et al., 2008;Chun et al., 2018).
Based on the author's initial observations through interviews with 100 civil servants in the Palu City government environment, 63 employees (63%) were dissatisfied with leadership in their institutions, and 55 employees (55%) stated that there were perceived injustices while working in their institutions.
Based on the results of the initial interview, which has been conducted as the preliminary study of this research, the organizational injustice felt by most civil servants will have an impact on the formation of negative emotions in the form of anger, shame, hostility, and the desire to retaliate (Barclay et al., 2005;Barclay & Skarlicki, 2009). The phenomenon is thus following research (Morris & Keltner, 2000;Cohen-Charash & Spector, 2001) that organizational justice can shape and influence individual emotions, such as delight, angry, and sad.
The gap discussed in this study relates to the results of research conducted by several researchers (Susanj & Jakopec, 2012;Gillet et al., 2013;Deschamps et al., 2016), which stated that there is a close relationship between leadership style and organizational justice. Transformational leadership has always been associated with results such as leadership effectiveness, satisfaction, innovation,quality improvement, performance evaluation both subjectively and objectively and organizational justice, even though the underlying process is not entirely clear (Bass, 1995).
There is still little research on organizational justice that considers emotions (Devonish et al., 2012;Barclay & Saldanha, 2015). Likewise, Fambrough and Hart (2008) and Rupp et al. (2014) stated that in the relationship between leadership and emotions, mediating or moderating variables need to be revealed.
Therefore, the novelty of this study is to explore the influence of leadership styles (transactional and transformational) on organizational justice (distributive justice and procedural justice) and their impact on the positive emotions of civil servants.It is because the perceptions of leadership style (transactional and transformational) and organizational justice (distributive and procedural) possessed by employees in those cities can vary, which positively can influence the formation of positive emotions. Bass (1996) defines transactional leadership as leadership that interacts with subordinates through the transaction process, as stated in social exchange theory (Bass & Riggio, 2006;Yukl, 2013). In transactional leadership, leaders will strive to create a climate conducive to the development of innovation and creativity (Bass & Riggio, 2006).

Hypothesis Development Transactional leadership
Transactional leadership is a type of leader that provides inspiration and intellectual stimulation to each of hisfollowers and has charisma towards followers (Yukl, 2013). The leader's task in this context is to try to motivate his subordinates to excel beyond previous expectations and forecasts so that his subordinates will receive rewards for his achievements (Northouse, 2016). There are three dimensions in this type of leadership, namely: Contingent reward, Active management by exception, and Passive management by exception (Bass, 1996).

Transformational Leadership
A transformational leader is someone who can bring about change in its members and the organization as a whole (Bass & Riggio, 2006). Transformational leaders try to bring each individual and team to work beyond the status quo and are leaders who have a vision going forward by making various changes in organizational culture and new vision values (Northouse, 2016). Bass and Avolio (1993) define transformational leadership is leadership that can change the behavior of subordinates into someone who feels capable and highly motivated and strives to achieve high and quality work performance. This type of leadership has four dimensions, namely: Idealized influence, Individualized consideration, Inspirational motivation, and Intellectualstimulation (Yukl, 2013).

Distributive Justice and Social Comparison
The first theory that uncovered psychological processes related to the formation of justice judgments is equity theory (Adams, 1966;Robbins & Judge, 2013). This theory provides the understanding that people judge justice by comparing the ratio of inputs they provide (for example, time, resources) are related to what they receive (salary, promotions, personal development opportunities), and then this ratio is also compared with the same ratio in others.
Distributive justice refers to perceived fairness regarding the organizational results that people receive. One of the basicprinciples in the distributive justice literature is that people form perceptions about fairness by comparing the results received with the results received by others, resulting in what is called social comparison (Kim et al., 2015).

Procedural Justice
In achieving procedural justice (Cropanzano et al., 2007;2008;Cropanzano & Stein, 2009) state that three main conditions must be met. First, in the procedure, the decisionmaking process consists of several people, not individuals. Second, the decision-making team has equal power among its members (Barclay & Skarlicki, 2009). Third, each team member involved in decision making must have the opportunity to get the same input (Barclay & Skarlicki, 2009).

Emotion
Emotion comes from Latin, which is emovere, which means "moving away" (Tracy et al., 2007). The meaning of this word implies that in the concept of emotions, the tendency to act is absolute. Emotions are reactions to stimuli from outside, and within individuals, for example, happy emotions encourage changes in one's mood, so physiologically seen laughing, sad emotions encourage someone to cry.

The Effect of Transformational Leadership on Procedural Justice
Transformational leadership is a type of leadership that involves empowering employees, giving individual consideration to subordinates, and supporting their ideas (Bass & Avolio, 1993). Transformational leaders enable employees to be able to influence the outcome of organizational decisions in their interests (Pillai et al., 1999).
Furthermore, transformational leaders motivate their followers or subordinates to engage in fair exchange relations (Pillai et al., 1999). Therefore, both of these factors constitute procedural justice, which in this case includes the extent to which aperson has the opportunity to vote in the decision-making process.
As Selznick pointed out that managerial authority only comes from employee agreement regarding psychological contracts, so they agree that their activities are managed by the management (Folger & Bies, 1989). Following what was stated by (Folger & Bies, 1989;Song et al., 2012;Kim & Kim, 2015) that transformational leadership is a leadership that focuses on the process of allocating organizational resources that aim to improve the development of employee skills. Thus, one crucial thing is managerial responsibility in implementing decision-making procedures that will guarantee employee or subordinate perceptions of procedural fairness (Pillai et al., 1999). H1: Transformational leadership has a positive effect on the perception of civil servants regarding procedural justice.

The Effect of Transactional Leadership on Distributive Justice
Bass stated that transactional leadership is based on theories regarding economic exchange (Pillai et al., 1999;Van Knippenberg et al., 2007;Gupta & Singh, 2015). Under transactional leaders, employees or subordinates tend to focus on fairness regarding the results received. Therefore, previous research (Konovsky & Pugh, 1994;Taxman & Gordon, 2009) state that distributive justice is a measure to assess the fairness of transactional contracts in economic exchange. Furthermore, this relationship can be related to the reality of one of the norms of distributive justice, namely that the parties involved provide mutual benefits in the hope that they will receive comparable benefits in the short term (Konovsky & Pugh, 1994;Taxman & Gordon, 2009).
In conjunction with performance evaluations, Greenberg and Baron (2000) argue that, if leaders are considered fair by their subordinates, the factors influencing subordinates' perceptions must be strengthened. It is done by ensuring that subordinates' expectations about the results they receive are closely related to their work (Pillai et al., 1999;Strom et al., 2014;Gupta & Singh, 2015).
Furthermore, previous research (Strom et al., 2014;Gupta & Singh, 2015) stated that until recently, research on the relationship between organizational justice and leadership tends to focus on transformational leadership. Transactional leadership also has a role in increasing the motivation and emotions of organizational employees (Strom et al., 2014;Gupta & Singh, 2015). When employees receive results that are following what they are doing (distributive justice) from their leaders, they will feel positive affections that have an impact on improving their performance.
As such, this is following transactional leadership. As stated by Bass that the function of the leader is to clarify the factors that influence subordinates' perceptions of fairness because they are closely related to rewarding their excellent performance (Pillai et al., 1999). H2: Transactional leadership has a positive effect on the perception of civil servants regarding distributive justice.

The Effect of Organizational Justice on Positive Emotions
When the employee is rewarded for his excellent job performance, receives a promotion, or receives gratitude from a colleague for helping the colleague, then the employee feels emotions in themselves. Thus, organizational justice is also an emotional experience (Cohen-Charash & Spector, 2001; Barclay et al., 2005;Cassar & Buttigieg, 2015).
Other studies that examine emotions (Ledimo, 2015;Budiyanti et al., 2018), state that procedural and interactional justice is justice that can predict individual emotions. Emotions are also able to mediate the relationship between perceived organizational justice and actions to take revenge (Cassar & Buttigieg, 2015). Violations in breaches of psychological contracts mediate the relationship between justice and organizational emotions (Cassar & Buttigieg, 2015). Perceived organizational justice about human resource management at work affects the quality of services provided by employees (Dzansi, 2016). There is a negative relationship between distributive and interpersonal justice with employee turnover (Moon, 2017). On the other hand, he also shows that there is a positive relationship between attributive, procedural, and interpersonal justice with positive emotions.
Emotions, including anger, happiness, and self-related emotions such as guilt, are used as dependent variables in multiple studies (Weiss et al., 1999;Budiyanti et al., 2018).
Based on the description, the hypotheses in this study are: H3: Procedural justice has a positive effect on the positive emotions of civil servants. H4: Distributive justice has a positive effect on the positive emotions of civil servants.
Thus, by the main objective of this study, namely to uncover the influence of leadership (transformational and transactional) on organizational justice (procedural justice and distributive justice) in explaining and predicting positive emotions of Government civil servants in the South Tangerang, Palu, and Sorong, the theoretical model proposed in this study is shown in Figure 1.

METHOD
In the quantitative stage, the authors conducted a questionnaire that would be used as a measurement of research based on the results of interviews that had been conducted with 25 (twenty-five) key informants. After compiling the questionnaire, the authors conducted face validity, social desirability response, and construct validity, which consists of convergent and discriminant validity. After the validity results obtained the minimum requirements that meet the expectations of the authors, the measurement of the effect of the leadership variable (transformational and transactional) on organizational justice (procedural and distributive), which subsequently affects positive emotions were commenced.

Measurement
In this study, measurement indicators are the development of previous studies conducted by (Bass & Riggio, 2006;Yukl, 2013). Transformational leadership consists of four dimensions, namely: idealized influence, which operational definition is leaders always put the interests of the organization above personal interests; intellectual stimulation is leaders always encourage subordinates to solve work problems rationally; individualized consideration is leaders always try to improve the self-development of his subordinates, and inspirational motivation is leaders are always optimistic in performing their duties and responsibilities. While transactional leadership (Bass & Riggio, 2006;Yukl, 2013;Northouse, 2016) operational definition is: Leaders provide an explanation of the rewards that subordinates will receive if they work well; distributive justice (Cohen-Charash & Spector, 2001;Colquitt et al., 2001), operational definition is: "To me, the rewards that I receive will reflect the effort that I put in the job. Procedural justice operational definition is the consistency of rules in the organization (Cohen-Charash & Spector, 2001;Colquitt et al., 2001). The last one is positive emotions (Morris & Keltner, 2000) which operational definition is "I am pleased with the leader's treatment of its staff ". The constructs in this study were measured using Likert-scale scales ranging from 1 to 7. 1 = Strongly Disagree, and 7 = Strongly Agree.

Instruments Testing
Instrument testing is carried out to test whether the research instrumentis powerful to measure the research construct in this study. A face validity, content validity, convergent validity, and discriminant validity (Hair et al., 2010) was also conducted to complete confirmed validity on every measurement.The pilot test in this study was conducted with the primary objective in testing Social Desirability Response (SDR) and construct testing (validity and reliability). The next step is to conduct the construct reliability testing. This test is carried out to test the consistency of the indicators used in research.

Sampling
The sampling in this study uses purposive sampling techniques aimed at civil servants in the city of South Tangerang, Palu, and Sorong with the following categories: (1) male or female, (2) civil servants, (3) willing to participate in the research. Sincethe formula for determining the sample size cannot be used for non-probability samples, the determination of non-probability sample sizes is usually based on the subjectivity of the researcher or comparison in previous studies (Hair et al., 2010).
The minimum sample size in research using SEM based on variance is 5 (five) to 10 (ten) times the number of indicators or ten times the number of parameters contained in the research model (Chin et al., 1995). In this study, there are four parameters (the number of arrows which are from the independent variable to the dependent variable), the minimum sample size for this study is 4 x 10 = 40 (forty) respondents (Chin et al., 1995).
The higher the sample size used, the better the results of the study because it will reduce the sampling error (Hair et al., 2014). Thus, the sample size in this study was set as many as 400 (four hundred) respondents using survey methods in collecting the data.

Data Analysis
The data analysis method used in this research is variance-based SEM with the help of PLS SMART Software. The main reason for using variance-based SEM in this study is because there are four constructs with formative indicators to be measured, namely transformational leadership, transactional leadership, procedural justice, and distributive justice. When wanting to measureleadership constructs, for example, transformational and justice constructs such as procedural using SEM, then these constructs must be treated as constructs with formative indicators (MacKenzie et al., 2005).
The statement of MacKenzie et al. (2005) is in line with Hair et al. (2011)SEM is equivalent to carrying out covariance-based SEM (CB-SEM, they argued that PLS-SEM could handle both formative and reflective measurement models. Reflective indicators are seen as functions of the latent construct, and changes in the latent construct are reflected in changes in the indicator (manifest) variables (Hair et al., 2011). SEM is equivalent to carrying out covariance-based SEM (CB-SEM). In contrast, formative indicators are assumed to cause a latentconstruct, and changes in the indicators determine changes in the value of the latent construct Hair et al. (2011) SEM is equivalent to carrying out covariance-based SEM (CB-SEM). Table 1 showed the characteristic of the respondents in this study. Based on Table 1, it can be seen that the significant characteristic of the respondent is Female, 41-45 years old, married, have been working for 1-10 years, are well-educated, most of them hold a bachelor's degree, and have monthly income of IDR 2,500,000-5,000,000.

RESULT AND DISCUSSION
While the Socially Desirable Response (SDR) test is conducted utilizing a non-paired sample, for each indicator of the five constructs measured. In doing so, researchers, assisted by two research assistant assistants, distributed questionnaires to civil servants living in Palu (30 people) and South Tangerang (30 people). For civil servants living in the city of Palu, questionnaires with direct questions were given, and civil servants living in the city of South Tangerang were We have given questionnaires with indirect questions.This test is carried out with nonparametric statistics using SPSS. For non-paired samples performed using the Mann Whitney test. The test results show that the value of p obtained more than 0.05, which means that the two samples (non-paired) come from populations that have the same average (mean) or expectations. In other words, the average respondent's answers from the two samples are the same. For more details, can be seen in Table 2.
The results of the calculation of the AVE value generated are 0.586 (more than 0.5) for positive emotional constructs so that it can be said, one construct in this study has a good convergence validity (Hair et al., 2014). AVE values that are only seen here are positive emotional constructs because they are constructs with reflective indicators. While the other four constructs, namely transformational leadership, transactional leadership, procedural justice, and distributive justice, are constructs with formative indicators, so the AVE value does not provide any information. Likewise, the construct reliability values seen in this study are only constructed with reflective indicators, namely positive emotions. Cronbach's alpha value generated was 0.811. The measurement of construct reliability in this study uses Cronbach Alpha (α), as stated by Nunnally (1978) andKaplanand Saccuzzo(1982), that a good α value for each construct used in basic research amounted to 0.70 to 0.80. Thus, the construct reliability used in this study is considered excellent. Table 3 showed that the results ofthe validity of the convergence and internal consistency of measurements for reflective constructs (positive emotions) in this study are reasonable. The AVE value indicates it for the construct, which is above 0.5, which shows that the validity of the convergence for the construct is good-likewise, the Cronbach alpha value and Composite Reliability for the construct in this study. Baumgartner and Homburg (1996) stated that Composite Reliability is considered better in estimating the internal consistency of a construct. Based on Table 4, the Composite Reliability value for this construct in this study is above 0.6.
To test the validity of formative constructs in this study, namely: distributive justice, pro- cedural justice, transformational leadership, and transactional leadership, a bootstrapping process is performed on the Smart PLS program, so that one of the outputs is produced, namely the outer weight output table. The author compares the value of T-statistics with T-tables (two-tailed). If there are indicators in each of these formative constructs that have a T-statistic value <1.96, then the indicator is not significant in the construct it constructs. Therefore, the construct is deemed not to meet the construct validity test criteria ( Jogiyanto & Abdillah, 2009). If the formative construct does not meet the Construct validity test criteria (i.e., there are one or more insignificant indicators), then the formative construct cannot be statistically tested further in the structural model. Table 4, which is the outer weight output, shows that the T-statistic value for formative construct indicators is > 1.96 (T-table). Thus, the formative construct in this research, namely; distributive justice, procedural justice, transformational leadership, and transactional leadership can be further tested in the structural model of this study.  Figure 2 explains the goodness of fit structural models in this study. In this study, the evaluation of structural model results begins with looking at the R-square value for each endogenous variable (positive emotion) in the model produced by Warp PLS. The R-Square value shows the ability of the independent variables in explaining the variance of the dependent variables. The results of this study indicate that the transformational leadership ability in explaining the variance of procedural justice is 81.30%; transactional leadership ability in explaining distributive justice variance is 0.790%, and the ability of procedural justice and distributive justice to explain positive emotional variance is 0.565%.
For the R-square value of 0.67, it is categorized as reliable, 0.33 is categorized as moderate, and 0.19 is categorized as weak. Thus, the overall research model shows that transformational leadership and transactional leadership have a strong ability to explain procedural justice and distributive justice, and procedural justice and distributive justice have a strong ability to explain positive emotions.
Furthermore, the value of f square (effect size) shows the effect of the independent latent variable on the latent dependent variable. Effect size values can be grouped into three categories (Kock, 2013;Hair et al., 2013), there are weak (0.02), moderate (0.15), and sturdy (0.35). The effect size value below 0.02 indicates that the influence of the independent latent variable is feeble from a practical perspective even though it has a significant p-value.The estimated results show the effect size of the effect of transformational leadership on procedural justice is 0.813; is 0.793, the effect size of the effect of procedural justice on positive emotions is 0.492, and the effect size of the influence of distributive justice on positive emotions is 0.575.
The effect size value of the influence of transformational leadership and transactional leadership on procedural justice and distributive justice is classified as a large effect size group. Likewise, the effect size of the effect of procedural justice and distributive justice on positive emotions is classified as a large effect size group. Thus, it shows that transformational leadership and transactional leadership have an essential role from a practical perspective in influencing procedural justice and distributive justice, as well as procedural justice and distributive justice for positive emotions. Q-square (usually also called the Stone-Geisser coefficient) is a non-parametric measure obtained through the blindfolding algorithm (Solihin & Ratmono, 2013). Q-square is used to assess the predictive validity or relevance of a set of predictor latent variables on the criterion latent variable. Models with predictive validity must have Q-square values greater than zero. The estimation results of this research model show good predictive validity because the Q-square value is above zero (Table 5).
Overall the Goodness of Fit (GoF) index of this research model is calculated based on the formula proposed by Tenenhaus et al. (2005). GoF = √(mean communality X mean R-square) notes: mean (average) Based on the results of this study the average value of communality is 0.734, and the average value of R-square is 0.722, then the Goodness of Fit (GoF) index of the research model is = 0.728 (Tenenhaus et al., 2005). A GoF value above 0.5 can be categorized well. Table 5 shows the results of SEM estimation and hypothesis testing. The table shows that all research hypotheses are supported because the value of Critical Ratio (CR) or t is statistically significant, and has a direction following the hypothesis, namely positive direction. The recommended CR value in SEM analysis to show that theright level of significance is more than 1.96 (Garson, 2016).
Based on the results of this study, that the leadership perception model canexplain and predict organizational justice and organizational justice can explain and predict the positive emotions of civil servants in Makassar, Palu, and Tangerang Selatan.
Hypothesis 1, which stated that transformational leadership has a positive effect on the perception of civil servants regarding procedural justice, is supported. The results of this study are generally consistent with the results of research conducted by Tyler (1986), which stated that perceptions of procedural fairness have a vital role in evaluating the leadership process, which in this case is transformational leadership. In this study, transformational leadership can explain and predict perceptions of procedural fairness at 81.30% (R 2 = 0.813). Hypothesis two, which stated that transactional leadership has a positive effect on the perception of civil servants regarding distributive justice, is supported, according to Bass(1985) research which stated that transactional leadership is a style of leadership that is based on the process of economic exchange while transformational leadership is based on the process of social exchange. Thus, under the subordinate transactional leadership, the focus is on his perception of distributive justice rather than procedural justice (Bass, 1985). Subordinates feel that their relationship with the leader is based on a transactional process, in which case the subordinate will receive some compensation for the work he does. Transactional leadership was able to explain and predict perceptions of distributive justice fairness at 79% (R 2 = 0.790).
Hypothesis three and four in this study, which are stated that organizational justice (distributive justice and procedural justice) have a positive effect on positive emotion, are supported. The results of this study are also generally following the results of Cohen-Charash and Spector (2001), which stated that perceptions of organizational justice could shape the emotions of employeesin an organization. In this study, the variables of the perception of fairness of civil servant organizations in Makassar, Palu, and South Tangerang were able to explain and predict positive emotions at 56.50% (R 2 = 0.565).
Indonesia public servant would see the fairness of the procedures, if they shall meet the following criteria, the extent to which they suppress bias, create consistent allocations, rely on accurate information, are correctable, represent the concerns of all the recipients, and are based on the prevailing moral and ethical standards (Budiyanti et al., 2018). As another aspect of the public servant's justice perception, procedural justice seems to act as an essential requirement. The violation of procedural fairness would not elicit positive emotions (Budiyanti et al., 2018). Distributive justice has also been found to be a significant determinant of positive emotions. As it deals with the perceived fairness of outcomes, it has the potential to have substantial implica-tions in the organizational context, of which distribution of outcomes is an integral part.

CONCLUSION AND RECOMMENDATION
The results of this study indicate that among civil servants, there is a significant relationship between transformational leadership and procedural justice, as well as transactional leadership and distributive justice. In this case, based on our literature review, there is still little empirical testing of these relationships. Hence, the results of this study support the view that transformational and transactional leadership can have a different impact on perceived organizational justice perceived by subordinatesbased on social and economic exchange relations. Economic exchanges are based on short-term transactions, but social exchanges occur because of the trust of individuals involved in exchanges regarding their respective obligations in the long run (Konovsky & Pugh, 1994). Transformational and transactional leadership differently are strictly related to perceptions of organizational justice.
The results of this study are also consistent with the research (Fambrough & Hart, 2008;Strom et al., 2014). They stated that one of the factors determining the formation of employee emotions is the rules that have been set. While the enforcement of these rules by the leader, whether it is following procedures or not.
Transformational leadership seems to influence procedural justice, which in turn forms positive emotions. It is because positive emotions are often something that results from extraordinary leadership, such as transformational, charismatic, and visionary leadership. Transactional leadership, on the other hand, seems only to affect distributive justice and also impact positive emotions. Researchers in the field of organizational justice has agreed and agreed that emotions have a significant role related to the occurrence of organizational injustice (Barclay et al., 2005). In general, employees in an organization would feel angry when given an inappropriate award and will feel guilty when over-valued (Homans, 1958). 180 There are several limitations in this research that need to be discussed. First, the data in this study were only collected from respondents in the public service sector, namely civil servants working in the government sector. Therefore, a generalization of results in other sectors in Indonesia cannot be made.
Second, this research also only focuses on the influence of leadership and justice perceptions at the individual level, not at the group level, when the workgroup is treated as a whole. Therefore, for further research, it is expected to be able to use the organization as a unit of analysis.
Third, the perceptions of justice are not formed in isolation, but in the context of specific relationships with several individuals and groups. Therefore, an assessment of an individual's perception of justice may be influenced by responses from other parties in an organization.