The role of budget decision making in mediating trust, social ties, cultural context and local government performance

Article history: Received: November 15, 2020 Received in revised format: January 28 2021 Accepted: March 2, 2021 Available online: March 2, 2021 Indonesian Corruption Watch (ICW) research reveals that corruption in Indonesia is mostly carried out by civil servants. An audit report from the Indonesian Supreme Audit Institution also finds local government internal control systems' weaknesses, especially controlling budget execution. This phenomenon is inseparable from the budget actors' decision-making process when preparing a budget. Budget decision making is a crucial process. Budget actors use their preference to choose among priorities and put the programs and activities which they have chosen into the budget. Social exchange such as trust and social ties can restrain budget actor distortion when making a budget decision and improve local government performance. Cultural context, in this research, refers to high context, affects individual behavior, and may influence the budget decision-making process. The result shows that decision making mediates trust and context culture to local government performance. Trust also proven can improve local government performance. Social ties, on the other side, do not have an influence both on decision making and local government performance. When trust has been intertwined among members, social ties no longer have a role in social exchange. Trust is the strongest form of social relation. Trust in this research is also formed because of the figure. A figure holds an important role in high context culture. This research pointed out that decision making by considering social values can make the decision-making process better and even improve local government performance.

Indonesian Corruption Watch (ICW) research reveals that corruption in Indonesia is mostly carried out by civil servants. An audit report from the Indonesian Supreme Audit Institution also finds local government internal control systems' weaknesses, especially controlling budget execution. This phenomenon is inseparable from the budget actors' decision-making process when preparing a budget. Budget decision making is a crucial process. Budget actors use their preference to choose among priorities and put the programs and activities which they have chosen into the budget. Social exchange such as trust and social ties can restrain budget actor distortion when making a budget decision and improve local government performance. Cultural context, in this research, refers to high context, affects individual behavior, and may influence the budget decision-making process. The result shows that decision making mediates trust and context culture to local government performance. Trust also proven can improve local government performance. Social ties, on the other side, do not have an influence both on decision making and local government performance. When trust has been intertwined among members, social ties no longer have a role in social exchange. Trust is the strongest form of social relation. Trust in this research is also formed because of the figure. A figure holds an important role in high context culture. This research pointed out that decision making by considering social values can make the decision-making process better and even improve local government performance.

Introduction
Research on local government budgets in Indonesia shows an interesting fact. The retardment of Local Government Budget approval is caused by a disagreement between the executive and legislative (Febrina & Isril, 2016;Parwati, Budiasih, & Astika, 2015;Subechan, Imam, & Haryono, 2014;Riyanto, 2012;Hanida, 2010). The results of the Indonesian Supreme Audit Institution also show that there are many irregularities committed by the local government, 46% of them are irregularities in the internal control system (BPK-RI, 2017). The Indonesian Corruption Watch (ICW) ranking results indicate that budget corruption cases in Indonesia are mostly done by civil servants. These findings indicate that local government performance needs to be improved. Budget decision making is a crucial process because people determine the achievement of performance since the planning process; of course, it sails from the decision-making process. Many regulations arrange the budget preparation mechanism, but budget decision-making is still a hard thing to do. Budget actors must consider the available resources. Limited resources force budget actors to make decisions using a priority scale. Priority scale setting is an effective method (Goodwin & Frew, 2013) but fraught with political intention (Onyango-Delewa, 2016;Amans, Mazars-Chapelon, & Villeseque-Dubus, 2015;Mutiganda, 2013;Ryu, Bowling, & Wright, 2008;Schick, 2007). Finally, budget actors burdened a variety of stakeholder interests (Cheung, 2006;Kelly, 2005;Fisher, Peffer, & Sprinkle, 2003).
Research related to budget decision making and performance provides interesting results. Research on Costa-Font, Forns & Sato (2015) revealed that social priorities and preferences are inherent in choosing programs and activities, individuals can change their views when they are involved in budgeting. The research of Ozdemir, Johnson & Whittington (2016) and Mutiganda (2013) found that ideology plays a role in the selection of program and activity preferences when making a budget decision. The results of previous studies have indicated that social preferences can play a role during the process of budget decision making. Social preferences can be triggered by a social exchange between budget actors within organizations. The social exchange contains a social value that can restrain the dysfunctional behavior of budget actors (Molm, 2014;Bottom, Holloway, Miller & Whitford, 2006;Lawler & Yoon, 1993;Cook & Whitmeyer, 1992;Cook & Emerson, 1978;Emerson, 1976). Social exchange theory is widely used in accounting research to explain various social phenomena that occur in organizations, such as the achievement of organizational and individual performance (Bal, Chiaburu & Jansen, 2010;Gould-Williams & Davies, 2005;Moran, 2005), as well as negotiations, social preferences and decision making (Pereira, Prada & Santos, 2016;Chuang & Schechter, 2015;Nye, 2009;Bottom, Holloway, Miller & Mislin, 2006;Bazerman, Curhan, Moore & Valley, 2000). Social exchanges can take various forms with the ultimate goal is the acceptable result to all parties. Over time, some researchers try to uncover the dark side of social exchange due to the imparity of social structure and power. Some researchers found the negative side of social exchange such as the emergence of restrictions on decision making (Wang, McNally & Lenihan, 2019), unethical compromise (Tangpong, Li & Hung, 2016), and the decline in organizational innovation (Molina-Morales & Martinez-Fernandez, 2009).
Trust is a form of social exchange that is inherent in the process of social relations. It takes a specific time to form trust (Fligstein & Dauter, 2007). Trust is not just a balanced reciprocal relationship. Trust can lead to a willingness to take risks, compromise, and even make sacrifices (Ertac & Gurdal, 2019;Lascaux, 2019). Social ties, on the other hand, is a form of accumulation of time, intimacy, and emotional factors. If the process runs well, it will form strong ties; on the contrary, if the process does not work well, it will shape weak ties. Trust and social ties are essential elements in making budget decisions that are full of political interests. In addition to the managerial implication of decision making, cultural context also plays an important role to explain the phenomenon. Organizations with high culture context tend to make decisions centrally. On the contrary, low context is more open to subordinate involvement and participation (Kittler, Rygl & Mackinnon, 2011;Kim, Pan & Park, 1998). The existing research of trust so far gave various results. Sward (2016); Buskens (2015); Pervan, Bove & Johnson (2009) stated that trust helps individuals reach an agreement and make a decision. On the other hand, excessive trust can lead to negative exchange relationships and reduce performance (Tangpong, Li & Hung, 2016;Bal, Chiaburu & Jansen 2010). The previous research of tie strength also brought a mixed result. Lawler, Thye & Yoon (2008); White & Houseman (2002), Friedkin (1982) has revealed that strong ties are better than weak ties in a social network. Liang, Liao & Liu (2017); Brown & Konrad (2001), Granovetter (1983) stated weak ties improve decision making and brought positive results for the organization.
The various results from the previous studies and private organizations' as a background context still open the opportunity to recent research to fill the gap. This study aims to determine the role of decision-making in mediate the relationship between trust, high/low context, and social ties to local government performance. Daerah Istimewa Yogyakarta (DIY) province was chosen as a research object for a couple of reasons. First, DIY won the title of the best local government for local government performance in 2019. Second, DIY has a particular characteristic that is not owned by other provinces, which is the governor, also a Sultan. Sultan is the highest leader in the area. Javanese culture considers the figure of the Sultan as a person who was elected, received revelation, has the substance manunggaling kawula gusti (the leader is the same as the people), and central decision making (Endraswara, 2013). This research contributes both theoretically and practically. Theoretical contributions related to the role of social exchange theory and culture can explain the process of decision making and performance in local governments. As for practical contributions, this research aims to reveal how trust, social ties, and power distance can improve decision-making and local government performance, and also keep budget actors to do the right thing.

Social exchange theory in the context of the budget
The theory of social exchange reveals that exchange is the basis of the social relations that exist between individuals and their environment. The individual is a rational creature who realizes that he cannot live alone and must relate to other parties for the sake of his survival (Boholm, Henning, & Krzyworzeka, 2013;Cook & Rice, 2003;Blau, 1977;Homans, 1958). As long as there is a social exchange within the organization, differences in interests cannot be avoided. The interwoven social network is certainly expected to run in a balanced way and produce comparable results (Cook, 2015;Charness & Rabin, 2002;Emerson, 1976). Social structure, on the one hand, can lead to imbalances in the social network. To achieve organizational goals and bridge the differences between negotiations, Lawler (1992) stated that bargaining is a form of cooperative decision-making of diverse motivations and alternatives. Negotiations cannot be separated from social structure and power, where more minority parties try to minimize the negative impact of negotiations (Tangpong, C., Li Jin., Ting Hung, 2016;Bottom, Holloway, Miller, Mislin, & Whitford, 2006;Cook & Emerson, 1978;Hamner, 1974). Social values become a counterweight that can create social balance (Cook, 2015;Cook & Rice, 2003;Cook, 1977;Emerson, 1976). These social values can be in the form of altruism (Zuckerman, Chen, & Nau, 2018;Murphy, Ackerman, & Handgraaf, 2011;Cropanzano & Mitchell, 2005), trust (Blay, Douthit, & Fulmer, 2018;Zuckerman, et.al, 2018), and fairness (Falk, Becker, Dohmen, Huffman, & Sunde, 2016;Cropanzano & Mitchell, 2005;Andreoni, Brown, & Vesterlund, 2002). Negotiations can be one way to strike a balance between two or more different interests (Ma, Yang, & Savani, 2019;Molm, 2014;Covaleski et al., 2003;Bazerman, et.al., 2000;Lawler & Yoon, 1993;Cook & Emerson, 1978). Honest behavior, commitment, openness, mutual trust between all elements of the organization can produce mutual exchanges (Blay, Douthit, & Fulmer, 2019;Fisher, Mitchell, Peffer, & Webb, 2019;Shea, Lee, Menon, & Im, 2019;Colquitt, Baer, Long, & Halvorsen-Ganepola, 2014;White & Klein, 2007;Cropanzano & Mitchell, 2005). Individuals tend to behave according to the preferences inherent in themselves (Jansson & Bursell, 2018;Levitt & List, 2007;Andreoni, Brown, & Vesterlund, 2002). In the context of the budget, budget actors may make the social exchange when making decisions. Preference arises from within an individual as a form of addressing social exchanges that occur. Preference refers to the main choices made by an individual through considerations that tend to contain fulfillment of individual needs. Therefore, when individuals are faced with social choices that are around them, then the individual has motivational considerations that ultimately drop the choices that provide incentives for him (Jansson & Bursell, 2018;Falk, et.al. 2016;Andreoni et.al. 2002). Preferences can appear during the process of selecting programs and activities that will be included in the budget. The existence of priority setting when the budgeting process can trigger budget actors to choose (Ozdemir, 2016;Costa-Font, et.al. 2015;and Mutiganda, 2013).

Local Government Performance
Schick (2014) states that performance is a case study of repeated failures. The involvement of all elements of the organization can improve performance (Wilfahrt, 2018;Kamau, Rotich & Anyango, 2017), clear and measurable performance indicators (Carlsson-Wall, Kraus & Messner, 2016;Halachmi, 2002), and ongoing monitoring and evaluation (Thiel & Leeuw, 2002). Soe & Drechsler (2018) viewed public performance from 3 (three) main aspects: service quality, outcome achievement, and community trust and participation. The government has issued regulations relating to local government performance. The government has issued Government Regulation No.3 of 2007; Government Regulation No.6 of 2008; and Government Regulation No.13 of 2019 issued to regulate the evaluation of local government performance. The regulation set the assessment of regional government policies, in general, includes implementation of technical policies, compliance with laws and regulations; and achievement of indicators and performance targets. It also includes transparency in regional financial management, the effectiveness of decision making; and compliance with minimum service standards.

Trust
Mcallister (1995) stated there are two types of trust, namely cognition-based trust (CBT) and affect-based trust (ABT). CBT is a trust that arises when someone, consciously and rationally, chooses to trust others. CBT arises from experience, level of social similarity, and consideration of organizational context (Mcallister, 1995). ABT, on the other hand, is a trust that arises as a result of emotional relationships between individuals. Haas & Deseran (1981) revealed that trust is a symbol of exchange that is not only based on utility exchange. Trust arises in various ways, like faith, gesture, and experience of social relationships (Sward, 2016;Buskens, 2015). Kollock (2002) citing Kelly and Thibaut's view where trust refers to a relationship that requires a willingness to take risks, compromise, and even make sacrifices (Ertac & Gurdal, 2019;Lascaux, 2019). Based on this understanding, excessive trust can lead to negative exchange relationships and can reduce performance (Bal, Chiaburu & Jansen 2010). It is essential to build trust in the realm of strategy and clear organizational rules (Tangpong, Li & Hung, 2016). In an organizational context, trust emerges gradually and requires a long process (Fligstein & Dauter, 2007). Trust in this study refers to previous studies on trust and organization (Wang, McNally & Lenihan, 2019;Colquitt, Piccolo, LePine & Zapata, 2012;Olson, Parayitam & Bao, 2007). Trust is a unique relationship and can have different results after the context. Trust between budget actors is an essential factor for reaching agreement and decision making (Blay, Douthit & Fulmer, 2019;Shea, Lee, Menon & Im, 2019;Wang et al. 2019;Colquitt, Baer, Long & Havorsen-Ganepola, 2014). Bottom, et.al (2006), Andreoni, Brown & Vesterlund (2002) stated that trust during the negotiation could improve the decision-making process and the performance of the regional government.
H 1a : Trust will positively influence budget decision making. H 1b : Trust will positively influence local government performance. 1058

Social ties
Social ties refer to a social relationship among members of an organization. Social ties arise from a combination of the amount of time in relationships, emotional intensity, intimacy, and reciprocal relationships that ultimately form the bond itself. We can categorize social ties into strong ties vs. weak ties. A strong bond can motivate an individual even when that individual in a weak position (Krackhardt, 1992). Strong ties can avoid misunderstandings. It is due to uniformity in understanding by members of the organization. The flow of information in organizations becomes better when there is a strong tie in it (Wilfahrt, 2018;White & Houseman, 2002;Friedkin, 1982). On the other hand, Breuer et al. (2013); Brown & Konrad (2001); Granovetter (1983) revealed that weak ties encourage organizational members to develop, open to changes, and improve sharing information across organizations. Weak social ties also prevent an individual from feeling a burden to return the favor that he/she ever received. We base this research on strong ties for several reasons: (1) strong ties can anticipate political resistance; (2) strong ties can facilitate changes that occur more positively. Budget decision making in a government context is fraught with change and uncertainty. Strong ties will have a positive impact on both decision making and achievement of local government performance.
H 2a : Social ties will positively influence budget decision making. H 2b : Social ties will positively influence local government performance.

Cultural context
The process of social exchange contains cultural elements because each individual carries the culture that he/she had acquired since birth. A culture that is inherent in a person can be due to the region of residence, ethnic background, race, and so on (Odongo, 2016;Berger & Luckmann, 1991). Hofstede (1980) dividing national culture into 4 (four) dimensions: small vs. large power distance, weak vs. strong uncertainty avoidance, individualism-collectivism, and masculinity-femininity. The culture developed by Hofstede has evolved and is widely used in organizational research. Gupta (2012) suggests that individuals must make adjustments related to culture that is inherent in themselves with the organization's culture to succeed in negotiations. Many researchers linked the cultural elements with the organizational context, especially in terms of decision making ( Wang et al., 2019;Vaisey & Valentino, 2018;Yates & de Oliveira, 2016;Gupta, 2012;Strengers, 2004;Vitell, Nwachukwu & Barnes, 1993). One cultural element often associated with decision making is the cultural context. Hall (1976) revealed the importance of cultural contexts in behavior. Hall divided context into high context and low context. High context refers to high hierarchy, group interest above individual interest, high involvement from the group member, and the flow of information is restricted and implicit. On the contrary, low context refers to low social structure, individual interest above group interest, the involvement of individuals is lower, and the flow of information is explicit (Gudykunst,1983). Gudykunst (1983) added that people in high context are more careful to interact with new people and screening new people from his/her background. People in low context are more open to strangers. People in high context are careful with intonation. They tend to keep a positive tone when they deliver their opinion or speak up their disagreement.
This research takes the object of working unit research in the environment of the Special Region of Yogyakarta. One of the unique features of Yogyakarta is that the governor is also a Sultan. Sultan is an aristocracy obtained from generation to generation and is the highest title of a leader. The Sultan's decree was an order that had to be carried out by the courtiers. Along with that, Javanese culture stressed the culture of ewuh-pakewuh (always reluctant), sendiko dalem (always say yes) to the leader (Endraswara, 2013).
H 3a : High/low culture will positively influence budget decision making. H 3b : High/low culture will positively influence local government performance.

Budget Decision Making
The budget decision-making process is a hidden social exchange. Individuals have an important role in the decision-making process because individuals choose and weigh the variables to be considered, individuals also supervise and determine the renewal of performance measures (Bazerman & Moore, 2009). Saaty (2008) revealed that decision making involves many unseen factors that sometimes we have to exchange. In general, individuals will make decisions that benefit themselves (Kamau, Rotich, & Anyango, 2017;Edwards, 1954). Pligt (2001) criticized Edwards' view and said that the decision-making process is not only seen from an economic perspective but also psychologically and socially. Decision making must consider the social aspects when decision making is collective which affects a group of people moreover the state (Frederick, Loewenstein, & O'Donoghue, 2002). Pligt (2001) dan Simon (1955) revealed that budget decision making contains social aspects and has an impact on government careers for budget actors (Frederick et al., 2002). A budget is a social contract with programs and activities included in the budget that aims at community welfare (Zaporozhets García-Valiñas, & Kurz, 2016;Bénabou, 2000). Fisher, Mitchell, Peffer, & Webb (2019) revealed that the insensitivity of budget actors when budgeting can have an impact on budget preparation that has a high error rate. Elangovan & Suddaby (2019) criticized the decision-making process that has existed, not considering the wisdom in decision making. Boholm, Henning & Krzyworzeka (2013) added that the decisionmaking process needs to consider cultural aspects. Decision-making indicators are based on the criteria used in Li, Wang, Huang & Bai's (2013) research and adapted to the context of budget decision-making in Indonesia. H3b

Method
This research is a quantitative study using survey methods. We employ the survey method because this research requires individual opinion data to see reality in the field (Groves, Flower, Couper, Lepkowski, Singer & Tourangeau, 2009). The survey method is appropriate for conducting behavioral research, answering unanswered questions, and explicitly providing valuable information for decision making (Yamanishi & Li, 2002). We developed the questionnaire by organizing a forum group discussion (FGD). FGD was attended by the budget compiler from four working units such as the Regional Development Planning Agency, Local Government Finance Office, Cooperatives, Small and Medium Enterprise, and Trade Office, and Inspectorate. The budget compiler provides information related to the conditions when the budget changes are made. The population of this research is the local government civil apparatus in the DIY Working Unit (DIY-WU). The sampling technique uses a purposive sampling technique with the following criteria: (1) respondents are involved in the DIY-WU budget preparation process; and (2) involved a minimum of one year in budget preparation. The data collection method was a survey pick-up method, in which the questionnaire was submitted and taken directly to the respondent after the respondent finished to fill the questionnaire. Data analysis using PLS-SEM with SMART-PLS 3 software to run the data. PLS-SEM was chosen due to some reason: (1) predicting and identifying key driver constructs, (2) achieves high levels of statistical power even with small sample sizes, and (3) can handles constructs measured with single and multi-item measures (Hair, Hult, Ringle & Sarstedt, 2017). Even PLS-SEM does not require normality, it does not mean we can ignore the data classification. Hair et al. (2017) stated that running data, the researcher needs to consider outlier data. An outlier that arose from primer data can come from respondents' inconsistency when filling the questionnaire. This problem can be fixed by running a boxplot test to identify the outlier data. After examining the data quality, we can run the PLS-SEM.

Result and Discussion
We distributed one hundred and ninety-five questionnaires to thirty-nine DIY-WU. There are one hundred and ten questionnaires that we can process from twenty-eight DIY-WU. The eleven rest, especially hospitals, cannot be reached due to pandemic. The first step before running the PLS-SEM is data examination (Hair, et al., 2017). Data examination was performed with a nonstatistic and statistical examination. A non-statistic examination was conducted by screening the completeness of the questionnaire. We ensured the respondents fill all the blanks in the questionnaire. Statistic examination was conducted by testing the outlier. An outlier is an extreme response to a particular or all questions. The boxplot test was carried out to identify the outliers and there was found twenty-two outlier data. The outlier data were excluded from the sample so the final data was eighty-eight data. The next step was running statistics descriptive to understand the demographic and the pattern of respondents' TR BDM HLC ST LGP 1060 answers. The demographic data were divided into six groups such as gender, age, position, length of holding the position, length of work, and level of education. The demographic data based on gender shown there is one hundred and seven men filled the questionnaire and the rest is female respondents. From the age group, there are five respondents who range age twenty until thirty, thirty-two respondents between thirty-one until forty, thirty-two respondents between forty-one until fifty, and nineteen respondents have age above fifty. The eighty-eight respondents are divided into four categories based on their position in the office. There are two heads of department, ten head of division/head of sector, fifty-three head of subdivision/head of sub-sector/head of section, and twenty-three functional planner. All of the respondents are actively involved in working unit budget decision making. Based on a length of hold position, there are forty-eight respondents hold the position one until three years, twenty-seven hold the position four until six years, nine respondents hold the position seven until nine years, and four respondents hold the position above nine years. Based on a length of work, there are twenty-eight respondents who have worked for one until ten years, twenty-eight respondents have worked for eleven until twenty years, twenty-eight respondents have worked for twenty-one until thirty years, and four respondents have worked for above thirty years. Based on the level of education, forty-two respondents are bachelor's degrees and forty-six respondents are graduate degrees. The overall demographic data showed although the majority respondent holds the position in the range one until three years, they already became an apparatus in quite a long time. Civil servants have a gradual and definite career, to hold some position they need passing through the stages as an ordinary apparatus. After data examination and running statistics descriptive to analyze the demographic respondents, the next step is running the data using SMART-PLS3.
The first stage of PLS-SEM testing is to test the validity and reliability. We can see the validity value of each variable from the outer loading value, where the outer loading value is 0.5-0.6 can still be accepted (Chin, 1998). We can also see the reliability value from the composite reliability value. Composite reliability has been used since it more appropriate to apply a different measure of internal consistency reliability. Cronbach's alpha in one way is sensitive to the number of items in the scale and generally tends to underestimate the internal consistency reliability. If the constructed value > 0.7, then it can be said that the construct has excellent reliability (Hair, Sarstedt, Hopkins & Kuppelwieser, 2014). Based on the test results, the outer loading value is above 0.5 and the reliability values of all variables are above 0.7. The result of outer loading and composite reliability values we can see in Table 1. After passing the validity and reliability test, the next step is testing the structural model. There are some important structural model metrics such as collinearity issues, R 2 , f 2 , Q 2 , and the significance of the structural path coefficients. Hair et al. (2017) stated when inner VIF values below the threshold of 5 than collinearity is not a critical issue in the structural model. All VIF values are below 5 so the model is freed from the collinearity issue. The R 2 value is the basis for understanding the model's predictive power. The R 2 value was 0,583 and included as moderate. The f 2 effect size is used to evaluate whether the omitted construct has a substantive impact on the endogenous construct. The f 2 effect size is categorized into three categories. Value of 0.02 represents small effects, 0.15 as a medium effect, and 0.35 as large effects. The larger effect was budget decision making to local government performance with 0.362, followed by a trust on decision making and trust to local government performance with 0.309 and 0.244 which is a medium effect, and the small and no effect was performed by high/low context to budget decision making with 0.099, high/low context to local government performance with 0.011, social ties to budget decision making with 0.009 and social ties to local government performance with 0.034. Q 2 values were used as predictive relevancy of the models. Q 2 values larger than zero indicate the path model's predictive relevance for a particular dependent construct. Q 2 values in this research were 0.228 for budget decision making and 0.408 for local government performance shows that all the endogenous latent variables were predictive relevance for the models. The hypothetical test results appear in Table 2 below: Based on the results of the trust variable, it was known that trust has a positive effect on budget decision making and local government performance. It shows that trust can improve budget decision making and performance. This result is in line with the social exchange theory, where trust arises as to the social relation increases. Based on his/her experience, individuals establish communication with others, evaluate the gestures, and choose to trust that person. There is a unique finding related to the trust variable. The indicator "All parties trust each other during the negotiation process" has the highest value. On the contrary, the indicator "no one takes advantage of each other" has the lowest value. This result indicates trust among apparatus in DIY provinces tend to affect-based trust (ABT), which arose from an emotional relationship (Mcallister,1995). Trust from emotional relationships drive individuals to compromise and relent to other parties (Ertac & Gurdal, 2019;Lascaux, 2019). Budget actors have long been the state civil service apparatus (AVE = 17 years) and have long been involved in the budget preparation process (AVE=4 years). This long relationship indeed has built mutual trust between DIY-WU fellow members. The next hypothetical testing is related to the role of social ties. Social ties have not a positive influence on budget decisionmaking and local government performance. One of the explanations that can be given is thorough to respondents' answers. The respondent's answer shows that the level of closeness has the lowest value. The coefficient value also shows a negative sign 1062 which means the ties are weak. These results contradict with Wilfahrt (2018); White & Houseman (2002); Krackhardt (1992) who stated that strong social ties can produce uniformity, knowledge transfer, and useful information acquisition. Another supporting answer is related to the respondents' demographic data. The majority of respondents are first-line managers followed by middle and top managers in local government, which the result mostly shape by subordinate perception than a superior point of view. The result of high/low context shows that high/low context has a positive effect on budget decision making. The positive sign means that the DIY province has a high culture. Kittler et al. (2011);Boyacigiller et al. (2004), Kim et al. (1998), Gudykunst (1983), Hall (1976) provides the insight that in high context, social hierarchy plays an important role, the individual's desires are limited by a high degree of self-control because the interests of the group are more important than personal interest. The flow of information is using restricted code and visible from act and behavior. The DIY province has strong Javanese cultural roots. The hierarchy of structure is very high, especially the history of the DIY province is the kingdom, in Indonesia, we called Kasultanan. A special characteristic of DIY is legitimate and recognized by The 1945 State Constitution of The Republic of Indonesia. The Sultanate of Yogyakarta also serves as governor of a DIY province. The characteristic of DIY and Javanese culture is embedded with ewuh pekewuh, sendiko dalem to the leaders. Respondents' answers support the theory where "ewuhpakewuh, sendiko dhawuh culture restrain all parties to communicate his/her disagreement to the leader" indicator has the highest outer loading value, followed by the "positive tone are important to keep the harmonization during the negotiation process" indicator.
The next research question that needs to be answered is the role of decision making as a mediating variable. According to Hair et al. (2017) there are three types of mediation: (1) complementary mediation; (2) competitive mediation; and (3) indirect-only mediation. The result shows that budget decision making is complementary mediation to the relationship between trust and budget decision making since the mediating sign and the sign of trust and budget decision making is the same. The budget decision making is indirect-only mediation for high/low context and budget decision making since only the indirect effect has a significant effect, and there is no mediation effect on social ties and local government performance relationship. Budget decision making indicator "Annual work plan is the result of collective agreements, not individuals" has the highest value from respondents. It aligned with the previous discussion of trust and high context which lead to high collectivism in making a decision. As for local government performance, it shows that all apparatus understands their obligation to achieve minimum community services.
Another interesting finding is the respondents' answers related to the achievement of the target. The achievement of the target has the lowest value. Respondents prioritize the achievement of minimum community services, followed by the implementation of technical policies and obedience to laws and regulations. This result may explain the achievement of DIY provinces related to winning the title of the best local government for local government performance in 2019. With no intention of setting aside the target achievement as important criteria for performance, this result gives us the insight that detailed planning and technical policy implementation brings better performance. The result supports Thiel & Leeuw (2002) argument that stated performance paradox can be prevented with different aspects of policy implementation, technical policy implementation can be one of that prevention.

Conclusion
Trust has proven to increase the quality of budget decision making and performance. Trust among apparatus in DIY province incline to the affect-based trust concept. The length of work can be a major cause of trust. If we looking from a high context point of view, trust can be formed because of the leader figure. Javanese culture believes that a leader is the representative of God on earth. It holds the philosophy that leaders have "sungai, prasaja, ajur-ajer"; "berbudi bawa leksana" nature, which is generous as the wind, giving to the people sincerely physically and spiritually, protect and have significant benefits for the people. These values live among DIY people especially the DIY province apparatus. These results give a new insight into how the leader has such high value and are obeyed by society. However, dependence only on one figure cannot last forever. The DIY province needs to strengthen its decision-making system, especially if we tracing back that indicator "annual work plan contains programs and activities that have benefits for the community" have the lowest value.
Social ties were not proven in this research. As we know from the respondent's answers that subordinates have a high involvement in the decision-making process just limited to their duties as subordinates. The level of closeness and communication intensity have a low value. It can be triggered by the high context in the DIY province. Javanese values of ewuhpakewuh (reluctant to express his/her opinion or disagreement), sendiko dalem (always say yes) to the leader widen the range between superior and subordinate, plus social hierarchy, the nobility title is significant in the social environment. Since the majority of respondents are subordinates, social ties do not affect both to budget decision making and performance. The negative sign between social ties and budget decision making and performance indicate weak ties among apparatus in DIY province, support the respondent's answers.
The DIY province is proven to have a high culture context. It is inseparable from a hereditary cultural heritage. The DIY province is one province that still holds tight to the root of Javanese culture in social life. The findings of this research enhance empirical evidence to the role of social exchange in improving decision making and organizational performance. The social exchange suggests that social values can support decision making wiser and have useful value for all interested parties. Culture, in context meaning, influences individual behavior and it is carried away in everyday life including work life. The values of a culture are given from generation to generation and obeyed by its members.
This study has several limitations. First, as the method of collecting data is a pick-up survey, we cannot guarantee respondents who fill in are the people we addressed. Second, the majority of respondents are subordinate so the results tend to be a representative subordinate point of view. Third, an in-depth interview may be needed to bring the insight of budget decision making in the working unit DIY province. Future research is very much expected to improve the validity of this research. Future research can fill the weaknesses such as increase the level of response rate by using an online questionnaire and balancing between superior and subordinate respondents to obtain balance judgment. Along with the important role of the cultural context in this research, future research can use another context that applies in the social environment.