Revenue from the taxation sector is the main pillar of revenue in the APBN that is used to meet government expenditure needs in the context of national development. The following graph describing the development of tax ratio of indonesia in the last five years
The graph above shows that in 2012 to 2016 the tax ratio experienced a drastic decline, which was only 10.3 percent, far lower than the previous year. The previous year, in 2012 and 2013, the tax ratio be at approximately at 11.9 percent and decreased in 2014 to 11.4 percent until 2016 was at the lowest level. In 2017 the tax ratio experienced a slight increase, but this figure is still relatively low.
According to Kogler, Muehlbacher, and Kirchler  tax compliance can be grouped into two, enforced compliance and voluntary compliance. Enfirce compliance that arises because of the allegation of coercion.While voluntary compliance that appears because because of the awareness of the taxpayer itself. Procedural justice, trust in authority and a norm as important factors tha in impove voluntary tax compliance [2,3,4,5,6]. Other studies test impact deterrence factors such as tax penalties, tax audit, and tax rate on tax compliance . Previous research show that deterrence factors tax sanction can impove tax compliance [8,9].However these results contradict other studies which state that in reality, the effectiveness of the sanctions system can reduce taxpayer compliance [10,11,12]
Research of Ratmono  show that tax penalties as deterrence factor has not impact on tax compliance. The inconsistency of the results becomes an interesting thing to be re-tested by developing voluntary tax compliance research using trust in authorities and moral norms as mediating variables. Van Dijke and Verbon  found that trust is important when authorities focus on justice for increase voluntary compliance. Moral norms are psychological-social variables that are thought to influence voluntary tax compliance. Torgler  found that tax morale possessed by someone determines the occurrence of tax compliance behavior.
Based on the explain above, this study examines the effect of detterence factor tax sanction and social psychological factors procedural justice, trust in authoritues and moral norms in affecting voluntary tax compliance. This study has a difference with previous studies  which uses deterrence factors tax penalties and tax audits and only uses trust as a moderating variable, this study uses a more complex social psychology variable which further explains the relationship between deterrence and psychology factors social with voluntary compliance. This study also uses trust in government authorities and moral norms as mediating variables. The next difference is the research method used, Ratmono and Cahyonowati  using an experiment, while this study uses survey methods.
2. Literatur Review
Teori slippery slope
One of the latest theories about tax compliance is the slippery slope model from . This theory states that social psychology and deterrence variables determine the level of tax compliance. Psychological-social variables tend to influence voluntary tax compliance while deterrence variables tend to influence tax compliance based on fear of negative consequences or forced tax compliance. Based on this slippery slope theory, the policy of increasing public trust in the tax authority must be prioritized in order to increase voluntary tax compliance
Effect of procedural justice on voluntary compliance
Procedural justice is the justice felt by taxpayers in the process of distributing tax rights and obligations whether it has been carried out according to procedures or not. Empirical studies show that if a person feels that the authorities carry out fair and existing procedures, they tend to ask for authority  and are more in line with the decisions taken by the tax authorities [16,6]. From the description above, procedural justice is thought to influence voluntary compliance. Following hypothesis is formulated:
Effect of procedural justice on trust in authorities
Fairness Heuristic Theory explains that people will observe carefully whether the tax authorities act procedurally fair. Several studies have revealed that when tax authorities follow procedural justice rules, such as refraining from self-interest and voting in the community decision-making process, assessing such a fairer procedure will increase taxpayers' trust [17,18]. Based on the description above, it is hypothesized
Effect of trust in authorities on voluntary complience
Trust emphasizes the relationship between taxpayers and tax authorities resulting from the trust of taxpayers in the actions of tax authorities. Wahl  found a positive impact derived from the strength and trust in paying taxes. The same findings are mentioned in the study by Muehlbacher, Kirchler, and  and , where they found trust from tax authorities to increase voluntary tax compliance. Based on the description above, it is hypothesized:
Effect of procedural justice on moral norms
Theory of moral reasoning in the context of tax compliance states that moral decisions are mainly influenced by the application of sanctions at a low level of moral reasoning, expectations of fairness at a moderate level, and the issue of justice at the highest level. Some studies  provide evidence that when taxpayers are treated fairly by the tax authorities, the treatment stimulates the internalization of collective norms (the process of personal norms). can be formulated the following hypothesis:
Effect of moral norms on voluntary tax compliance
Norm activation theory  relavan to explain the behavior of taxpayers in fulfilling their obligations to pay taxes. Someone will obey paying taxes on time, if someone already feels that paying taxes is his duty. Wenzel  proved that personal norms negatively affect tax avoidance which means that it can increase tax compliance. Based on this, the following hypothesis can be formulated
Effect of tax sanctions on trust in authorities
Moral assessment of punishable behavior is a key factor in understanding the relationship between tax authorities and compliance sanctions . Whether tax disobedience is seen as immoral or immoral depends on how sanctions are perceived: (1) as compensatory that is to show an economic transaction, or (2) retributive shows that norm violations deserve punishment. Based on the description above, it is hypothesized
Effect of tax sanctions on Moral Norm
Tax sanctions are imposed on taxpayers as a legal consequence if they violate existing regulations. The theory that is considered to be closely related to the moral of taxpayers is the theory of moral reasoning. The theory of moral reasoning in the context of tax compliance states that moral decisions are mainly influenced by the application of sanctions on low moral reasoning.Based on the description above, it is hypothesized
Trust in authorities mediated reslationship procedural justice with voluntary compliance
The Heuristic Justice Theory explains when tax authorities act according to fair procedures, the taxpayer will increase taxpayer trust [17,18]. Trust in authority has a positive impact on the willingness to pay taxes. This is supported by  and , where they found trust from tax authorities to increase voluntary tax compliance
Moral Norm mediated relationship procedural justice with voluntary compliance
Fair procedures can affect the attitude of taxpayers to behave honestly and be responsible for their tax obligations. Thus a fair procedure will increase the morale of the taxpayer. Verboon and van Dijke  research proves that the moral norms of taxpayers have a positive effect on tax compliance. Wenzel  proves that personal norms negatively affect tax avoidance which means that it can increase tax compliance. Based on this, the following hypotheses can be formulated:
Trust in authorities mediated relationship tax sanction with voluntary compliance
Moral judgments about punishable behavior are key factors in understanding the relationship between imposing sanctions by tax authorities and tax compliance . Retributive sanctions tend to increase trust compared to compensatory sanctions because they show moral disagreement with norm violations . Thus the amount of sanctions that are felt to be fair will increase trust so that it will increase voluntary compliance. Based on the description above, it is hypothesized:
Moral norms mediated relationship tax sanctions with voluntary compliance
Theory of moral reasoning states that moral decisions are mainly influenced by the application of sanctions at a low level of moral reasoning. Moral norms owned by taxpayers can be intrinsic motivation for taxpayers in complying with their tax obligations without coercion . Based on the description above, it is hypothesized:
3. Research Method
Population and sample
The population in this study is an individual tax payer who has a business registered at Tax Office Partama Tampan which amounts to 8,832 (source: tax office Pratama Tampan). The sampling technique used convenience sampling, which is sampling in the easiest way according to the researcher. The selection of this sampling technique is used because the population in the study is homogeneous
Voluntary tax compliance measured by an instrument developed from  which consists of 10 question items that measure voluntary self-registration, refuse to register not according to procedures, pay taxes without coercion, awareness of paying taxes for the future period, feel paid tax is an obligation that should, understand taxation regulations, pay taxes with self-awareness, desire to contribute to the country and others.
Procedural justice measured by instruments from [26,24] with 5 items of questions that measure the same treatment with other taxpayers, correct errors in existing procedures, can give opinions in the tax authorities, sanctions have been given according to procedures, procedures are in accordance with moral ethics
Sanctions consist of four questions that measure the magnitude of sanctions, namely the amount of criminal sanctions, administration, sanctions as a tool for education, sanctions without tolerance.
Trust in government authorities
Trust in ttax authority measured by the instrument from  consisting of 5 items of questions that measure trust in government transparency, fairness in fulfilling obligations, equal rights of citizens, ability to manage taxes, have high integrity, act on the interests of citizens.
Moral norms are individual norms that are owned by a taxpayer, but may not be owned by other taxpayers. The instrument consists of 5 statement items that measure attitudes to the obligation to pay taxes, crimes if not paying taxes, not paying taxes is an act that is not good, the obligation to pay taxes even though it is unfair, paying taxes is an obligation to help the country.
All variables are measured using a 5-point Likert scale, point 1 = strongly disagree up to 5 = strongly agree.
From the 120 questionnaires distributed, 100 questionnaires could be accepted for data analysis. Table I show the descriptive data:
The data shows that the standard deviation is much smaller than the mean which reflects the representative data.
The hypothesis testing of the research was carried out by using the Structural Equation Model (SEM) approach by using the Partial Least Square (PLS) software. The results of the analysis can be explain as follows:
Result of outer model
Result of common method Bias
The result of common method biased by looking at the full collinearity VIF value of each variable can be seen in table II
In table 2 above can be seen the full collinearity VIF value of all constructs (first Order Construct) has the value of "Full Collinearity. VIF below 3.3. It can be concluded that it is free from collinearity or biased common method.
Result of test validity
Convergent validity is used to measure the same construct. Convergence is shown by high correlation. Test results of convergent validity can be seen in the table below:
The result of convergent validity analysis shows that all indicators have a loading value above 0.5 which means that all indicators are valid.
Result of discriminant validity
To see the results of the discriminant validity test by comparing the value of the correlation indicators in the construct with other variables, the correlation can be seen in the cross loading value in table 3 above. Discriminant validity can also be seen from the square root value AVE in the table 4 below
We can be seen from the square root value AVE greater than the correlation between latent construct variables. The square root of AVE procedural justice is 0.754, tax sanction 0.768, trust 0.721, moral norm 0.792, voluntary compliance 0.831 is greater than the correlation value between constructs so the above values are said to be good and meet the requirements of discriminant validity.
Result of test relialbility
Reliability test results can be seen in table IV above. Composite reliability values above 0.7. It can be concluded that all construct variables have high reliability which means that each construct variable has an average correlation between items in a reliable model test
Result of inner model
To predict the existence of causality in SEM-PLS by using Warp PLS 6.0 program. Figure of Full structural Equation Model is below:
R-squared only exists for endogenous variables  The results of the R-squared value can be seen in table IV above
The construct variable trust influenced by procedural justice and tax sanction of 0.106 (10%) moral norms (NM) has R square of 0.067 which means that it is only affected by procedural justice and tax sanction of 6.7% and voluntary compliance has a R Square of 0.067 which means it is affected by trust, procedural justice and moral norms are only 6.7% and the remaining 93.3% are influenced by variables or factors outside this research model.
Result of hypothesis
The results of testing hypotheses can be seen in the table below:
Table 5 shows the results of the testing of hypothesis1 is the effect of procedural justice on voluntary compliance has not significant (p= 0.156 0.05). The results of testing of hypothesis 2 shows p value = 0.003 is smaller than the critical value 0.05 which means that hypothesis 2 is accepted. The results of testing of hypothesis 3 shows p value= 0.091.0.050.05, which means that hypothesis 3 is rejected, The results of testing hypothesis 4 shows the influence of procedural justice on moral norms has a p value = 0.014 smaller than critical value 0.05, which means that hypothesis 4 is accepted. The results of testing hypothesis 5 shows that effect moral norms on voluntary compliance has a p value = 0.013 smaller than critical value 0.05 which means that hypothesis 5 is accepted. The results of testing hypothesis 6 shows the p value = 0.03 0.05, which means that hypothesis 8 is accepted, The results of testing hypothesis 7 shows the p value = 0.057 0.05, which means the hypothesis is rejected.
Effect of the mediating variable trust on the dependent variable voluntary compliance does not significant with p value = 0.091 0.05 which means that hypothesis 8 is rejected. The effect of tax sanctions on trust as mediation variables has a significant effect with p = 0.03 0.05, while the effect of trust on voluntary compliance is not significant with p value = 0.092 0.05. Both of these relationships are not significant so that the hypothesis 9 rejected. The result show that the trust in authorities not mediated the effect of procedural justice on voluntary compliance.
The results of testing hypothesis 10 shows the effect of procedural justice on moral norms is significant with p value = 0.014 0.05, while the effect of moral norms on voluntary compliance is also significant with p value =0.013 0.05. Both significant relationships so that the hypothesis 10 accepted. It can be concluded that moral norm mediates the effect of procedural justice on voluntary compliance. The results of this study are consistent with the research [32,20,1] found a positive impact derived from strengths and beliefs about tax payments.
The results of testing hypotheses 11 show that the effect of tax sanctions on the mediating variables of moral norms is not significant with p value = 0.057 0.05, while the influence of moral norms on voluntary compliance is significant with p value = 0.013 l 0.05.
The results of testing of hypothesis showing no significant influence of procedural justice on voluntary compliance. The results of this study are not consistent with previous research (Faisal, 2014 and Faizal, 2017) The result shows that procedural justice has positif significant effect on trust in government authorities. The result consistent with previous research [17,18] wich show the higher perceived procedural justice will increase trust in government authorities. The results shows that trust in the government has no effect on voluntary compliance. The results show that trust in the government does not affect the desire of taxpayers to pay taxesThe results of this study are not in consistent with Schwarzenberger [28,1].
In this study procedural justice effect has significant effect on moral norms. The correlation shows a negative value of -0.211. This indicates procedural justice actually decreases morality. The results of the study support the theory and consistent with previous studies [29,2,30] have succeeded in proving that the procedural justice of tax authorities has a positive effect on one's norms.
This study shows that moral norms effect on voluntary compliance. Correlation results show a value of -0.215 which has a negative relationship which indicates that increasing moral norms actually reduce voluntary compliance. Although this hypothesis is supported but this negative relationship is different from previous research. Basri and Al Azhar  research shows that increasing morale will reduce unethical behaviors such as tax fraud.
Tax sanctions has significant effect on trust in government authorities. The higher sanctions can increase the taxpayer's morality so that they are aware of their tax obligations. Consistent  that sanctions in the form of imposing penalties are perceived as indicators that tax authorities can increase honest taxpayer trust. The test results show that tax sanctions do not affect moral norms. The results of this study are not consistent with Novia (2016) that the existence of sanctions will affect morale.
Mediation testing used partial step method. This study shows trust in authorities cannot mediate the influence of procedural justice on voluntary compliance. The results of this study are inconsistent with the research  who found trust as a variable that mediates the effect of procedural justice on tax compliance.The results also shows that the trust in authorities not mediated the effect of procedural justice on voluntary compliance.
This study shows that moral norm mediates the effect of procedural justice on voluntary compliance. The results of this study are consistent with the research [32,20,1] found a positive impact derived from strengths and beliefs about tax payments.
The results show that the effect of tax sanctions on the mediating variables of moral norms is not significant effect. This study does not support the Theory of moral reasoning states that moral decisions are mainly influenced by the application of sanctions and  which states sanctions will increase moral obligations
The results show that there is no significant influence of procedural justice on voluntary compliance. Procedural justice has a significant influence on trust in government authorities. The higher perceived procedural justice will increase trust in government authorities. In this study, trust in the government does not affect voluntary compliance, but procedural justice affects moral norms. Moral norms affect voluntary compliance. The results show that the higher moral norms actually reduce voluntary tax compliance. Tax sanctions affect trust in government authorities. The higher sanctions can increase the taxpayer's morality so they are aware of their tax obligations. The results of the study show that tax sanctions have no effect on moral norms, meaning that trust cannot mediate the influence of procedural justice on voluntary compliance. trust does not mediate the influence of procedural justice on voluntary compliance trust does not mediate mediating the influence of procedural justice on voluntary compliance. The results show that Normal mediates the influence of procedural justice on voluntary compliance, but moral norms do not mediate tax sanctions with voluntary tax compliance.
This research is only carried out on individual taxpayers who have a MSME business only. The study area was only conducted in the areas of Pekanbaru, Kampar and Bengkalis so that these results cannot be generalized more broadly. Deterrence and social variables are still few and have not shown a large influence on voluntary compliance. The research model only uses mediation of moral norms and trust in government authorities
8. Suggested for Future Research
Further research should expand the object of research not only on individual taxpayers but also corporate taxpayers. Further research can expand the research area so that research results can be more generalized. Subsequent research can examine deterrence and social variables other factors social norms, economic levels, law enforcement, examination, politics etc. Research can further develop the model by adding other variables to obtain the best model in voluntary compliance
This study has a contribution to the theory that supports that procedural justice influences trust and moral norms, as well as tax sanctions also affect procedural justice and moral norms. Moral norms are intermediaries for the creation of voluntary compliance. It is important for the government to increase the taxpayer's morality so as to increase voluntary compliance.