Tax Authority Versus Peer Communication: The Influence of Trust, Service Climate, and Voluntary Cooperation

INTRODUCTION Trust can be built through effective communication because communication plays a vital role in shaping consumer perceptions (Ball et al., 2004). (Liu & Horsley, 2007) showed that government communication would build closer relations with citizens. Trust to the government will also emerge if the information provided, either directly or through the websites, is well conveyed (Schaupp, Carter, & McBride, 2010). Trust can arise from good communication patterns, especially if the information obtained from the government is accurate (Kogler, Muehlbacher, & Kirchler, 2013). A good pattern of communication between the two parties will ultimately have an impact on increasing voluntary cooperation (Kogler et al., 2013). (Hofmann, Beverungen, Räckers, & Becker, 2013) stated that we need to concern the communication between the government and citizens and also mainly on how the community receives and responds to the ABSTRACT

tool makes it easy for us to interact, exchange information with anyone (Wang, et al., 2012). Peer communication plays a role as a factor that influences the level of trust in a product or service (Wang et al., 2012). Peer communication is assumed to be a factor in the occurrence of norms that have a particular role (Wartick & Rupert, 2010). Someone will tend to think "what other people do," including peers or individuals, will follow what most people do (Wenzel, 2004). In general, people will often make decisions or choices based on the people around them (Bursztyn, et al., 2014). The role of peers as friends of discussion will provide an honest review of something discussed, and inevitably people will be more likely to believe or be influenced by peers (Yu, et al., 2004). Peer communication be the essential reference group to form a communication pattern because they trust each other and even share the social identity and have interests with the same references (Cotterell, 2007). They can share their identities and information because they believe that the information they share with others will be safe and secure (Ge, Figueiredo, Jaiswal, et al., 2003). (Bursztyn et al., 2014) showed that peer communication also plays a vital role in influencing other friends to decide whether to invest in the financial market.
In this study, we used two generations, namely generation X (born in 1965 to 1979) and millennial (1980 to 2000) (Jurney, et al., 2017) because they are the largest generation of taxpayers . Researchers assume that the most suitable system to maximize tax revenue for both generations is through voluntary cooperation. These two generations also proved to prefer the application of hard power and soft power in a balanced manner by the tax authority in Indonesia . Hard power and soft power in combination might be perceived as legitimate expert power , increase trust by creating the impression that free riders will be penalized while supporting honest taxpayers in order to achieve high-level tax compliance (Hofmann, et al., 2014). Based on the current conditions in Indonesia, it is tough to rely solely on enforced compliance systems because the number of tax officers compares with taxpayers in Indonesia is 1: 7700 (Misbakhun, 2018).
The contribution of this study is the first study that examines the influence of tax authority communication on trust, perception of service climate, and voluntary cooperation. Through communication, the tax authority can fortify or prevent the circulation of bad news about the tax institution. Through massive communication, the level of trust, perception of service climate, and voluntary cooperation continue to increase. Second, this research is the first to compare the effect of tax authority communication and peer communication on trust, perception of service climate, and voluntary cooperation. We found that tax authority communication had a more significant influence on trust, service climate, and voluntary cooperation than peer communication.
Theory of normative social behavior (NSBT) developed by Rimal & Real (2005) and also used in research Real & Rimal, 2007) proves that the existence of interaction or communication influenced individual behavior .  uses NSBT to see whether there is an influence of peer communication. It is necessary to know first normative social influences on decisions and risk behavior to explain in more detail the role of peer communication concerning normative social influences (Real & Rimal, 2007). These two types of norms can be referred to as norms about what has done compared to what norms should be done (Chung & Rimal, 2016;Rimal & Lapinski, 2015), which influences individual behavior through interaction in groups. Communication between peers is a substantial factor in determining normative social behavior (Southwell & Yzer, 2007). These normative factors may have a direct impact on the formation of someone's behavior (Jang, et al., 2013). Siahaan (2012) defines communication as a way or activity to convey information through the exchange of ideas, messages, or information such as speech, writing, or actions and behavior. Excellent communication will always be associated with the quality and quantity of the information delivered (Siahaan, 2012). The research of (Liu & Horsley, 2007) shows that communication built by the government will build closer relations with citizens. For example, Twitter's social media can be used to form relationships between tax authorities and taxpayers on an ongoing basis through tax socialization and opening questions on Twitter accounts. The government communication that delivered either directly or through websites can be successful if the government can build relationships through giving the information that needed and give the answer to the questions of taxpayers (Schaupp et al., 2010).
DGT communication via social media is regulated in the Tax Law, the Law of Public Services, and the regulation of Minister of Finance (table 1). The Act and the regulation are explained that the tax authority communication is carried out with the aim of improving public services, providing legal certainty, and increasing trust in public services. This law was prepared with the aim of providing justice, improving services to taxpayers, increasing certainty, and law enforcement.

Act of Republic of Indonesia number 25 the year 2009 concerning Public Services
This law was prepared with the aim of the tax authorities being obliged to serve every citizen to fulfill tax rights and obligations and to build public trust in public services.

Regulation of Minister of Finance number 174 the year 2012
This law was prepared with the aim of improving the organization and work procedure of the information office and complaints services of the Director General of Taxes (DGT).
The main task of government authorities is to report decisions and actions through the media to offset negative information about the government that citizens receive in the media (Liu & Horsley, 2007). This communication will build a very close relationship with citizens (Liu & Horsley, 2007) so that it will raise a high level of trust. Trust can arise from good communication patterns, especially if information obtained from the government is accurate, and will directly affect the trust of the government and will also have an impact on individual compliance (Kogler et al., 2013). Based on the explanation, the first hypothesis is H 1 : Tax authority communication affects trust The tax authority believes that taxpayers are citizens who are responsible for carrying out their obligations, then taxpayers also see the tax authority follow their group (Alm, 2014). Someone will tend to think "what others do," which means individuals will follow what most other people do (Wenzel, 2004). Peer influences individuals in making decisions (Real & Rimal, 2007), especially the decision about trusting the tax authority. Based on the explanation, the hypothesis is H 6 : Peer communication affects trust The results of media analysis have shown that peer communication has a vital role in the process of forming social norms of society, and influences behavior and actions (Wartick & Rupert, 2010).  uses the theory of normative social behavior (NTSB) to see the influence of peer communication. Understanding the role of peer communication concerning the risk of norm development does not only contribute to the theoretical understanding of normative social influences (Southwell & Yzer, 2009) but also communication that occurs between peers is a substantial factor in the formation of norms and behavior (Southwell & Yzer, 2007). Coercion from the government is not enough to increase an individual's tax, but rather an environmental factor in which a person will grow and have a doctrine about taxes (Rothengatter, 2005). One of the environments is peers. (Rothengatter, 2005) argues that the environment or peer has a significant role in contributing to the individual level of tax compliance. H 7 : Peer communication affects voluntary cooperation In service climate, the tax authority is considered a professional institution and performs the rules well, by providing services to its clients or taxpayers (Alm & Torgler, 2011;Braithwaite, 2017). So that the condition will improve compliance (Farrar & Thorne, 2016). Besides, taxpayers give trust to the tax authorities in the hope that the tax authorities can carry out their duties under the rules so that taxpayers pay taxes voluntarily (Kirchler, et al., 2008). Trust in tax authority gives a good perception of service climate and finally increase the voluntary cooperation of taxpayers (Gangl et al., 2016). Trust in the tax authority has a significant influence on service climate as well as on voluntary cooperation (Gangl et al., 2015). Based on the literature and studies described earlier, the researcher sets the hypothesis as follows: H 8 : Peer communication affects service climate as a partner who can honestly manage tax returns. Trust in the tax authority will also be high because taxpayers consider tax authorities to carry out their duties properly according to applicable rules (Gangl et al., 2016). Taxpayers not only follow existing laws but also have high enthusiasm to obey the law and see tax payments as a moral obligation (Braithwaite, 2017). High and low level of trust of the tax authority determines the perception of taxpayers concerning service climate (Gangl et al., 2012). Trust in the tax authority has a significant influence on the perception of service climate (Gangl et al., 2015). Based on the explanation, the hypothesis are H 2 : Tax authority communication affects service climate H 3 : Tax authority communication affects service climate through trust as a mediating variable (Hofmann et al., 2013) in their research stated that the most important thing is not only communication between the government and citizens. However, it is essential to give attention how the community receives and responds to the communication. The information contained in government communication will also increase the cooperation and compliance of taxpayers (Farrar & Thorne, 2016). Thus our study sets the hypothesis based on the above explanation: H 4 : Tax authority communication affects voluntary cooperation H 5 : Tax authority communication affects voluntary cooperation through service climate as a mediating variable (Yong & Martin, 2017) gave statement that the influence of peer communication is inseparable from how an individual makes decisions. For example, Asian people like to gather and tend to have at least a community, so that the level of decision making affects the group where the person is located (Yong & Martin, 2017). In contrast to western culture, which tends to be individualistic, westerns may only be influenced by their families who tend to have their perceptions and beliefs in seeing things (Yong & Martin, 2017). In this case, what we need to know is that a person does not always take a paradigm, selfish, and rational nature but instead follows the situation or group they are in (Yong & Martin, 2017). It means that the level of individual adherence will be influenced by the group as long as they believe that there are social penalties that they get if they do not

RESEARCH METHOD
Respondents in this study are individual taxpayers who were self-employed in two business fields, namely retail/production and services/ professions, which are domiciled in the cities of Jakarta, Surabaya, Semarang, Bandung, and Denpasar. The research sample was 120 respondents. The sampling technique used quota sampling with the following criteria: (1) the respondent is an registered taxpayer, (2) the number of respondents who own a business in the retail/ production or services/profession are equal (3) the number of male and female respondents are equal, (4) respondents represent X and Millennial generation, (5) respondents used social media to communicate about taxes and used tax authority services (email/live chat/twitter/telephone), (6) respondents represent the method of "bookkeeping and recording" to calculate income tax, bookkeeping and recording methods are determined based on business circulation in a year which is higher than 4.8 billion or smaller than 4.8 billion Rupiah.
There are several reasons for using the six criteria of respondents. Criteria number four aims to get the right respondent where the respondent use social media and is an active taxpayer when the survey is conducted. Criteria number two is used with the aim that respondents surveyed are only selfemployed taxpayers who use the self-assessment system to calculate their own taxes so that they are relevant to voluntary cooperation in this study. Criteria number three and six aim to reduce the bias of research results (Jurney et al., 2017).
The measurement used for each variable in this study is a seven-point Likert scale. The use of a seven-point scale in this survey aims to distinguish slightly favorable, moderately favorable, and extremely favorable feelings towards an object. According to (Krosnick & Presser, 2010), the use of a seven-point scale is appropriate. The Likert scale is a proper measurement for this study because it directs respondents to be able to give an opinion about agreeing or not to a question (Cooper & Schindler, 2014). Variables related to this study are peer communication, tax authority communication, trust, service climate, and voluntary cooperation, which are measured by a Likert scale from strongly disagree to strongly agree.
This study measure peers communication using three indicators, direct persuasion, state norms, Based on the identification of variables and hypotheses, the research model can be illustrated in Figure 1. My friend encouraged me to pay and report taxes correctly PC2 My friend encourages to report all income in the tax return PC3 My friend in the social media group suggested obeying to pay taxes PC4 My friend and I discussed the positive impact of tax on the country in social media PC5 I read in the social media group about the benefits of paying taxes AC1 Email service and ",kring pajak" (live chat, twitter, and telephone) answer all the obstacles that I experienced AC2 Communication with tax employees is easier and more convenient with email services and "kring pajak" (live chat, Twitter and telephone) AC3 The latest tax rules information through newspapers and magazines, greatly help my tax obligations AC4 Tax seminars and training held by the Directorate General of Taxes effectively helped my obstacles TR1 I rely entirely on the information on the website of the Director General of Taxes (www.pajak.go.id) to answer my taxation problems TR2 I believe in the tax authority because I agree with their vision TR3 I trust the tax authorities because they work competently and professionally TR4 I trust the tax authorities because they complete the task well SC1 The relations between tax authorities and taxpayers such as companies and clients SC2 The relationship between tax authorities and taxpayers is service oriented SC3 When I pay taxes, I do it because the Directorate General of Taxes might respond to my cooperative actions VC1 When I pay taxes, I do it because the Directorate General of Taxes treats me properly as long as I admit my mistake VC2 When I pay taxes, I do it because the Directorate General of Taxes helps taxpayers who make unintentional mistakes VC3 I pay taxes because it is easier than use tricks againts the tax authorities Source: Processed Data, 2019

RESULTS AND DISCUSSION
The questionnaires were distributed in three stages. The first step was determining the criteria for respondent validity included in the survey sample, which is taxpayers, self-employment in the retail/ production business or services/profession, aged 19 -54 years (Millennials and X generations), used social media to discuss tax with peers, and communicated with tax services (email/live chat/twitter/telephone). Respondents who did not meet these criteria were excluded from testing. The second stage was determined by the quota sampling criteria based on gender and income tax calculation methods. The respondent tested must represent both the male and female groups, as well as the "bookkeeping" and "recording" method group. The third stage of questionnaire distribution was carried out based on convenience sampling in a face-to-face and online form. The total number of respondents who responded was 164 respondents. Of the total questionnaires responded, only 120 respondents (73.17%) met the criteria and could be further processed.
highlighting positive aspects of paying tax (Onu & Oats, 2018). We modified the questionnaire (Wang et al., 2012) for tax authority communication. Authority communication indicators use several types of communication used by tax authority institutions, namely online communication, written communication, and face to face communication. We combined and modified questions of previous research (Alghamdi & Rahim, 2016;Wang et al., 2012). For trust variables, researchers used three indicators, relevant goals, competence & benevolent, and supported. Our study combined and modified the questions from (McAllister, 1995) and (Gangl et al., 2015) concerning the trust. We use the perception of service orientation as an indicator of service climate (Gangl, et al., 2015). Voluntary co-operation uses questions modified from (Gangl et al., 2015). The questionnaire is in the Table 2.  Table 4. shows the average value of all indicators used. The highest mean in peer communication (PC) variables is PC3 (5.27 out of 7) with the indicator of state norms. This shows that in communication between peers on social media, topics about the norm of paying taxes is the most discussed. Meanwhile, for authority communication (AC), the highest mean is on the question of AC2 (5.2 points out of 7) that the communication with the tax authorities through live chat/twitter/ telephone is more comfortable and more convenient, according to the respondents.  Table 5. shows that all indicators met the validity requirements with a loading value of > 0.50. According to (Kock & Lynn, 2012), the loading value and cross-loading that meet the convergent validity standard are in the range of -1 to 1, and the value in brackets must be higher than the cross-loading value. Convergent validity is acceptable if the Loading value is above 0.5, and the P-value is below 0.05 (Hair, Black, Babin, & Anderson, 2014). Cross loading must be smaller than the loading value.   Table 6 presents the test values, which reveal that the discriminant validity is fulfilled for all variables with the correlation coefficient value smaller than the AVE square value. The composite reliability must be more than 0.7 (Kock & Lynn, 2012). The results of the composite reliability of all variables are more than 0.7 (table 7) so that it is acceptable. All Cronbach's alpha values are acceptable when each value is above 0.5 (A. M. Dall'Oglio et al., 2010;I. Dall'Oglio et al., 2015;Nguyen et al., 2019) Table 9 shows the direct and indirect effects between variables. The direct link between AC and TR shows a p-value <0.01 with  value 0.440, which shows a significant positive link, meaning H1 is accepted. Direct links between AC and SC also show significant positive results (H2 accepted) with pvalue 0.015 and  value 0.192. The indirect effect between AC, TR, and SC shows a significant positive link with p-value <0.001 and  value 0.210 or H3 accepted. The direct link between AC and VC also shows a significant positive result (p-value <0.001 and  value 0.319, meaning H4 is accepted. The indirect effect between AC, SC, and VC shows a significant positive link with p-value 0.097 (significant at  10%) and  value 0.210, meaning H5 is accepted. The direct link between PC and TR; PC and SC also shows significant positive, so that H6 and H8 are accepted. The test results show that the direct link between PC and VC is not significant, and indirect links between PC, TR, SC, and PC, SC, VC also are not significant. However, H7, H9, and H10 are rejected. We found that the effects of authority communication to the trust, service climate, and voluntary compliance, showed strong significant effects (Table 9). It explained that Millennials and X generations make an individual assessment on every information they get from all resources, including the internet, social media, and print media. Millennials and X taxpayers considered all the tax information coming from DGT as ways to help, simplify, and get closer to taxpayers so that taxpayers can perform their obligations in a simple way. This was the duty of tax authority as a legitimate expert power in the view of Millennials and X generations (Tjondro, Setiabudi, et al., 2019). This also confirms the H3 test results, where authority communication affects the perception of service climate through trust. The tax authority is more trusted by the Millenials and X generations because they get reliable information and served as clients. This trust is formed as a result of quality reviews or discussions and certainly helps in various ways, such as providing the information and helping in solving problems (Schaupp et al., 2010;B.G. Southwell & Yzer, 2007;Wang et al., 2012).
On the contrary, based on the result, it can be concluded that peer communication only had a significant effect on perceptions of service climate, but low influence on trust and does not affect the voluntary compliance (table 10). It explained that all the presumptions about peer communication as a factor in the occurrence of norms (Wartick & Rupert, 2010), as a reference in making decisions and choices, and as a reference of discussions to provide honest reviews (Yu et al., 2004), only has a strong effect on perception of service climate and weak effect on trust of DGT. This result describes the overall picture of Millennial and X taxpayers that they do not rely on their decisions about tax institutions on several individual peers. institutions, perception of service climate, and v o l u n t a r y c oo p e ra t i o n r a t he r t h an p ee r communication on all three. Table 10 shows the authority communication and peer communication to have a significant effect on the trust of the tax authority institution. Nevertheless, the value of  between AC and TR is higher than PC and TR, meaning authority communication has a more significant influence on trust than the effect of peer communication on the trust of the tax institution. We also found authority communication has a stronger influence on climate service perceptions than peer communication on the perception of service climate. The influence of authority communication on voluntary cooperation proved to be higher than peer communication because peer communication proved not to affect voluntary cooperation. comfortable and more convenient and time-effective. The second-highest average indicator is AC3 question (5.13 out of 7), which is about print media (newspapers and magazines). The taxpayers are much helped by disseminating the latest tax rules through print media so that the taxpayers can have an information update.
Based on Table 11. The AC2 indicator has the highest average than others (5.2 out of 7). This shows that the taxpayers agree concerning authority communication with taxpayers through online communication (e-mail, live chat, telephone, and Twitter). This means that taxpayers, especially the X generation and Millennial generation, really enjoy and like this way of communication because it is more

CONCLUSION
Our research found that authority is more powerful in effecting trust and perception of service climate of tax institution, and voluntary cooperation rather than peer communication. This means taxpayers in this era, Millennials, and X generations, have more trust in authority communication than peer communication in terms of taxation. Image of the tax institution has a more direct influence on trust, perception of service climate, and voluntary cooperation in terms of taxpayers' assessment of the tax institutions. This is an advantage that needs to be utilized by the Directorate General of Taxes (DGT).
We also found that taxpayers prefer tax authority communication through online communication (e-mail, live chat, telephone, and Twitter) and print media (newspapers and magazines) than other regular media. This means that taxpayers, especially the Millennial and X generation, really enjoy and like this way of communication because it is more comfortable, more convenient and time-effective. Millennials and generation X is the most significant tax contributors to the country. Therefore, their needs based on technology must be considered including in authority communication.
This study has several limitations. This study uses non-probability sampling, which is quota sampling, which has weaknesses in determining samples that tend to be subjective for researchers because they are carried out in the number or quota desired by researchers. The results of this study apply in the location sector that we input as the research target, but it must be noted that it cannot be used in general (depending on where the research was conducted). For better research results for further research in population, sampling needs to be considered further.