Tax Compliance After the Implementation of Tax Amnesty in Indonesia

The Slippery Slope Framework argues that taxpayers’ trust in government and their perceptions of government’s power explain the extent of their tax compliance. This study empirically examines such influences on taxpayers’ perceptions of tax compliance (either voluntary or enforced) in Indonesia after the implementation of a tax amnesty. In 2018, data were collected through a face-to-face questionnaire survey of 410 small business taxpayers and analyzed using a stepwise linear regression model. Trust in government was positively related to agreement with/support of voluntary tax compliance; however, perceptions of government’s power were only marginally significant. Neither trust in government nor perceptions of government’s power influenced taxpayers’ perceptions of enforced tax compliance. These findings partly support the Slippery Slope Framework. This study contributes to the literature on the influence of trust and power on voluntary or enforced tax compliance in developing countries, particularly Asia, after the implementation of a tax amnesty.

compliance here only because we consider Indonesia alone and no change was made to the country's taxation system during the observed period.
Between July 2016 and March 2017, Indonesia implemented a tax amnesty. One of the reasons leading the government to take this step was to account for the investments and withdrawals of Indonesian and foreign funds that were not yet disclosed in its calculations (Idris, 2016). The tax amnesty successfully increased the amount of tax collected because the total amount of declared, previously undisclosed income was valued at IDR 4,813 trillion (US$361.5 billion), and approximately IDR 135 trillion (US$ 10.2 billion) in penalties were collected (Ministry of Communication and Informatics, 2017). The large amount of revenue collected in penalties indicates that the extent of tax compliance had been low.
In developing countries such as Indonesia, there are numerous reasons why tax noncompliance may be high before the implementation of a tax amnesty program. First, taxpayers' lack of accurate tax information may cause them to pay less than the complete amount of their taxes (Kirchler 968793S Park & Hyun, 2003), particularly on assets that cumulate every year. Moreover, the combination of complex tax laws and a lack of understanding of taxes significantly influence tax compliance (Eriksen & Fallan, 1996;Inasius, 2019a;Kirchler et al., 2008;Torgler & Schaltegger, 2005). Second, for most taxpayers, the moral dimension of tax compliance may be weak, and many may seek ways to avoid paying taxes (Alm & Torgler, 2011). These taxpayers have no intention of paying the complete amount of their taxes despite having an adequate understanding of the tax laws. However, the inconveniences of using undeclared income make tax amnesty a powerful measure to increase tax compliance among both honest and dishonest taxpayers. A tax amnesty program increases people's likelihood of declaring income with special rates and exempts those who did not accurately report their past incomes from legal prosecution (Alm et al., 2009;Baer & Le Borgne, 2008).
However, most of the studies applying the Slippery Slope Framework were conducted in developed countries (e.g., Kastlunger et al., 2013;Kogler et al., 2015;Lisi, 2012;Muehlbacher et al., 2011;Prinz et al., 2014). Inasius (2019b) formalized the Slippery Slope Framework of tax compliance in a developing country, whereas Batrancea and colleagues (2019) conducted studies in both developed and developing countries before the implementation of tax amnesty; the empirical findings reveal that trust is an important factor influencing voluntary compliance. In contrast, power had no significant influence on either voluntary or enforced tax compliance. This study employs the Slippery Slope Framework (Kirchler et al., 2008) to analyze perceptions of tax compliance and focuses on the relationship between taxpayers' perception of trust in the government and perceptions of government's power to induce tax compliance (provide punishment for evasion, e.g., fines) after the implementation of Indonesia's tax amnesty.
Research on taxpayer behaviors and taxpayers' perceptions after the implementation of a tax amnesty program in developing countries might identify changes in tax compliance. Before the implementation of tax amnesty, tax evaders who want to be honest taxpayers may worry that their lack of disclosure will be detected by the government and result in penalties; this fear may be lessened after tax amnesty implementation, which may cause them to declare their untaxed incomes. Therefore, increasing tax compliance by enhancing the government's power might be relatively effective before implementing a tax amnesty. The question is whether taxpayers' tax compliance will increase if the government focuses on enforcing compliance after a tax amnesty. Therefore, a better understanding of taxpayers' motivations following amnesty situations can inform and improve tax compliance policy and perhaps increase the effectiveness of tax policies. Since taxpayers may have both economic and noneconomic reasons for their decisions, tax amnesty programs may be more effective if tax authorities consider appeals to noneconomic motivations, such as people's trust in authorities.
In particular, earlier research, such as that by Torgler and Schaltegger (2005), has explicitly considered the underlying reasons for taxpayers' reactions to tax compliance following a tax amnesty disclosure. This study contributes to the tax compliance literature by demonstrating the importance of trust in encouraging tax compliance in developing countries such as Indonesia. Furthermore, this study provides empirical evidence that power may also influence voluntary compliance, although this effect is weaker in Indonesia than in developed countries. The Indonesian tax amnesty is considered one of the most successful worldwide, as it collected the largest amount of revenue in Asia (Ibrahim et al., 2017), and Indonesia has cultural similarities with several other Asian countries. Hence, this study contributes to the development of an Asian perspective on tax compliance by highlighting the determinants of tax compliance after the implementation of a tax amnesty. The results will assist Indonesian policymakers in ensuring voluntary and enforced tax compliance while formulating tax laws and regulations.
The remainder of this article is organized as follows. First, a brief review of the literature is provided. Subsequently, the study's methodology is described, which is followed by a report of relevant statistical results. Finally, the findings are discussed in the "Results" section, and conclusions are drawn.

Literature Review
Tax compliance is defined as the accurate reporting of income and expenses according to applicable tax laws (Alm, 1991;Slemrod, 2007). It indicates taxpayers' willingness to obey tax laws and their extent of obedience to such laws to maintain their country's economic equilibrium (Andreoni et al., 1998). Following Alm (1991), Andreoni et al. (1998) broadly defined tax compliance as taxpayers' willingness to comply with tax laws regarding the accurate declaration of income, filing of returns with the appropriate government agency, and payment of all taxes in a timely manner.
Tax amnesty is an option chosen by many governments to increase the tax base and tax revenues (Alm et al., 2009;Dunn et al., 2016;Torgler & Schaltegger, 2005). This fiscal policy provides a set period during which taxpayers can pay a fixed amount of taxes in exchange for forgiveness of their tax liability without fear of legal prosecution (Alm et al., 2009;Baer & Le Borgne, 2008). Under Indonesia's tax amnesty law, people who owed taxes were pardoned and not subjected to financial or criminal sanctions if they paid a fixed amount of taxes rather than the complete amount of unpaid taxes. The proportion of unpaid tax was 0.5% of the undeclared assets from June 2016 to March 2017; for nonsmall and medium-sized enterprises (non-SMEs), the proportions of unpaid tax were 2% of the undeclared local assets and repatriation of unpaid foreign assets for July to September 2016, 3% for October to December 2016, and 5% for January to March 2017. Furthermore, the foreign assets' declarations were 4% for July to September 2016, 6% for October to December 2016, and 10% for January to March 2017 (Ministry of Finance of Republic of Indonesia, 2016).
Most of the earlier studies on the relationship between tax amnesty and revenue were conducted in the United States (Alm & Beck, 1993;Luitel & Sobel, 2007). Torgler and Schaltegger (2005) conducted research in two other countries to assess whether people agreed with the implementation of a tax amnesty and found that individuals were less likely to agree than disagree with a tax amnesty, but tax amnesties tended to increase tax compliance (Dunn et al., 2016;Luitel & Tosun, 2013;Torgler & Schaltegger, 2005).
If an amnesty is not followed by an increased cost of compliance, or if the cost of enforcement is too high, the goals of the amnesty may not be realized (Bose & Jetter, 2012;Dunn et al., 2016). Tax amnesty offers opportunities for tax evaders to avoid severe penalties on discovery, and they tend to disclose their undeclared income during the amnesty if they think that their tax evasion will otherwise be detected by the government (Alm & Beck, 1993;Alm & Torgler, 2011;Dunn et al., 2016). Alm et al. (2009) argued that a tax amnesty will improve tax compliance if programs are implemented to induce many taxpayers to participate in the amnesty, the government's tax services are improved, and better education on taxpayer responsibilities is provided and stricter sanctions are applied on tax evaders after the tax amnesty. However, critics contend that the actual experiences of many countries indicate that the immediate impact of amnesties on revenues is usually quite small (Alm et al., 2009). Therefore, if most taxpayers do not voluntarily comply with tax laws, they might interpret tax amnesty for the most significant instances of tax evasion as the government's support of social equality. If honest taxpayers disapprove of special treatment for certain tax evaders, then their compliance might decline.
There are two ways to understand tax compliance in the literature on the development of tax compliance: the economic perspective, which is based on the economic theory, and the behavioral perspective, which involves sociological and psychological aspects (Alm et al., 2016;James et al., 2001;Kirchler et al., 2010Kirchler et al., , 2014. Tax rates, penalty rates, and the likelihood of being audited are economic factors, whereas individual perceptions of social equality, fairness, and justice along with social norms are behavioral factors (Kirchler, 2007;Kirchler et al., 2014;Kogler et al., 2015;Slemrod, 2016;Torgler & Schneider, 2005). The economic approach is derived from the economic approach to criminality proposed by Becker (1968). Allingham and Sandmo (1972) applied this perspective to tax compliance by developing a model in which noncompliant taxpayers choose to evade taxes when that option offers the highest expected gains. However, subsequent empirical studies have had mixed results regarding predictors' influences on some aspects of compliance, such as tax rates, the likelihood of an audit, and the severity of fines (Kirchler et al., 2008). A further criticism of the economic approach is that laws are limited to incentive-oriented behaviors, which might not be the only solution to noncompliance (Frey, 1997;Kirchler et al., 2014). Therefore, thorough psychological analyses of tax behaviors are required to understand tax compliance and evasion.
The Slippery Slope Framework (Kirchler et al., 2008) has wide acceptance and is cited by many earlier studies on tax compliance (Andyarini et al., 2019;Batrancea et al., 2019;Inasius, 2019b;Kogler et al., 2013Kogler et al., , 2015Lisi, 2012;Mas'ud et al., 2019;Muehlbacher et al., 2011;Prinz et al., 2014). In this framework, voluntary and enforced compliances are predicted by indicators of authoritative power and measures of trust between taxpayers and the government. These two variables and their interrelationship are used to predict the extent to which taxpayers voluntarily comply or are forced to comply with tax laws. Therefore, the Slippery Slope Framework differentiates between synergistic and antagonistic taxation environments. A mutually respectful and trusting relationship between taxpayers and the government characterizes a synergistic environment, and it is argued that transparent government procedures that respect and support taxpayers' interests encourage voluntary compliance (Kirchler et al., 2008). However, when taxpayers and the government lack mutual trust, an antagonistic environment in which citizens pay taxes only because they must, which reflects the assumed need for and uses of enforced tax compliance, is created (Slemrod, 2016).
It is important to clarify the difference between enforced and voluntary tax compliances (Braithwaite, 2007(Braithwaite, , 2009). Citizens perceive their governments to be powerful when compliance is enforced through the detection of tax evasion. When audits are believed to be frequent and effective, the government's power is considered strong. Mutual trust is created in a context of respect and motivation to preserve the relationship (Kirchler et al., 2008). Tyler (2006) pointed out that taxpayers would likely comply if they believed that the government was legitimate. In this type of environment, there is little need for expensive enforcement mechanisms because, in the atmosphere of mutual trust, voluntary compliance is likely.
Over the past 50 years, tax amnesty has been offered by both developed and developing countries in response to economic problems (Baer & Le Borgne, 2008;Torgler & Schaltegger, 2005). Although earlier empirical studies show that tax amnesty affects the amount of tax revenues collected and increases the tax base (Alm, 1991;Bose & Jetter, 2012;Torgler & Schaltegger, 2005), empirical evidence from various countries that have conducted tax amnesty in the last few decades reveals that the benefits of tax amnesty are not significant (Alm, 2019;Alm & Beck, 1993). Therefore, a better understanding of taxpayers' motives to participate in tax amnesty can improve the effectiveness of tax amnesty programs (Dunn et al., 2016) and, in turn, improve subsequent tax compliance.

Units of Analysis and Respondents
In Indonesia, SMEs are clearly defined regarding adherence to tax rules. Sales revenue is one criterion used to determine the tax laws for SMEs in the country (Inasius, 2015). Indonesia's Ministry of Co-operatives and SMEs (2008) defined the laws based on turnover. A micro enterprise has a turnover (net sales) less than IDR 300 million (US$21,614), a small enterprise has a turnover less than IDR 2.5 billion (US$180,115), and a medium enterprise has a turnover less than IDR 50 billion (US$3,602,305). In this study, SMEs were considered the units of analysis.
The study's target population of Indonesian taxpayers comprised retail SMEs who participated in tax amnesty. Among the 966,000 taxpayers who participated in the tax amnesty (Alm, 2019), 431,000 were categorized as SMEs countrywide. The study selected SME data because this category has a lower level of compliance than large businesses (Susila & Pope, 2012). Among the 60 million SMEs in Indonesia, only approximately 1.5 million have reported and paid their taxes (Inasius, 2019b). Furthermore, Jakarta was selected as the target population area because it is the center of government, as well as the center of Indonesia's economy and culture. In addition, retailers in Jakarta generally act as role models for retailers operating in various other regions across the country.
To obtain an analysis sample, 1,000 self-employed individual retailers (self-employed taxpayers) and small enterprise retailers (as defined by the 2008 law) were selected using purposive random sampling to obtain a sample of Indonesian taxpayers with an annual turnover less than IDR 50 billion (equivalent to US$3,602,305 at the time of the study). All the cases in the sample were SMEs operating in either traditional markets or shopping centers. A survey was conducted on self-employed individual retailers in Jakarta Province, where most of the economic activities take place and the largest share of Indonesia's tax revenue is concentrated (Inasius, 2019a). The data were collected in April through June 2018 through face-to-face interviews. The survey covered five municipalities: Central Jakarta, West Jakarta, North Jakarta, East Jakarta, and South Jakarta. The random selection of respondents, distribution of questionnaires, and collection of data were conducted by 54 supervised students. The respondents' participation in the study was voluntary, and they were informed that their answers would be kept confidential. The respondents completed the questionnaires with the students' help in approximately 30 min. As an incentive to participate, potential respondents were offered credit points that could be used for a free 30-min consultation to obtain answers to tax-related questions. The respondents filled out the questionnaires at their places of business. Before filling out the questionnaire, the respondent was given the opportunity to view the questions. The questions that were not answered by the respondents were categorized as missing data. Among the 1,000 administered questionnaires, 578 were rejected because of missing data (the respondents did not answer whether they had participated in tax amnesty) and 12 were excluded as they were not correctly completed. Hence, the final sample comprised 410 completed questionnaires; the response rate was 41%.

Data and Variables
The questionnaire was adapted from the study by Muehlbacher et al. (2011) by carefully translating it from the original German into English and from English into Indonesian and then back-translated into English to verify translation accuracy. Both the Indonesian and English versions were provided to the respondents. The analysis included four constructs (Trust, Power, Voluntary Compliance, and Enforced Compliance), each consisting of three questionnaire items. The two dependent variables were Voluntary Compliance and Enforced Compliance. For the 12 items, responses were denoted on Likert-type scales, with options ranging from 1 = completely disagree through 5 = completely agree. Any "Don't know" response was considered missing data. For Cronbach's alpha, the lower limit of acceptability is .6 (Sekaran & Bougie, 2011); the range for each item was .63 to .84, which signified that a measure was consistent and reliable (Hair et al., 2010;Sekaran & Bougie, 2011). Table 1 depicts the means, standard deviations, and reliability statistics for the four constructs and their items. The items for each variable were analyzed by calculating the mean and standard deviation. Voluntary Compliance had the highest mean (3.69); it was followed by Trust (3.67), Power (3.56), and Enforced Compliance (3.12).
The questionnaire collected data on gender, age, educational attainment, and income, which served as the analysis' control variables. Educational attainment was categorized into three groups: compulsory (equivalent to junior high school), A-level (equivalent to senior high school), and academic (equivalent to having a bachelor's degree). Furthermore, three income categories were considered: low (annual turnover not exceeding IDR 2.5 billion [US$180,115]), medium (annual turnover not exceeding IDR 4.8 billion [US$345,821]), and high (annual turnover not exceeding IDR 50 billion [US$3,602,305]). Table  2 depicts the survey participants' demographic data.

Methods of Analysis
Two stepwise linear regression analyses were conducted to assess the influences of Trust and Power on attitudes toward tax compliance. The model assessed the relationships between independent and dependent variables (Sekaran & Bougie, 2011). All the variables were standardized before conducting the regression analyses. First, the influences of Trust and Power on Voluntary Compliance were assessed; subsequently, estimates of their influences on Enforced Compliance were made.
Step 1 included Trust, Power, and an interaction term between Trust and Power, and Step 2 added the control variables to the model. Table 3 depicts the results of the regression of Voluntary Compliance. In Step 1, Trust was a statistically significant positive predictor of perception of Voluntary Compliance (p < .01) and, although weaker, Power was a positive predictor (p < .10), as well. The inclusion of control variables in

Voluntary Tax Compliance
Step 2 slightly changed the results. Among the control variables, age alone had a statistically significant effect (p < .01), which was positive. It meant that support for nonvoluntary compliance increased with age. Furthermore, regardless of gender, age, or socioeconomic status, those with high levels of trust in the government were more likely than those with low levels of trust to support voluntary tax compliance. However, Power had only a minor influence, and the effect of Trust did not depend on the perception of Power (or vice versa) according to the nonsignificant interaction effect.

Enforced Tax Compliance
The regression model used to test the effects on Enforced Compliance was identical to that used to test the effects on Voluntary Compliance. Table 4 reveals that Trust had a negative effect on Enforced Compliance, although the coefficient was weak (p < .10). Neither Power nor the interaction term was statistically significant in Step 1 or Step 2. Furthermore, adding the control variables in Step 2 had little impact on the results of Step 1. However, educational attainment and gender positively influenced Enforced Compliance, indicating that agreement with Enforced Compliance was higher among the more educated than the less educated respondents. In summary, Trust had a slight influence on the agreement with Enforced Compliance, which indicated that perceptions of a strong trust in government were associated with less agreement with Enforced Compliance. However, the influence was weak, and it disappeared in Step 2 when control variables were included in the model. Furthermore, men were more likely than women and respondents with higher education were more likely than those with lower education to agree with the Enforced Compliance items.

Discussion and Conclusion
As proposed by the Slippery Slope Framework, trust in government significantly influenced Voluntary Compliance.
Trust had a small negative influence on Enforced Compliance; however, it lost its significance once control variables were included. The perception that the government has taxation power was more important to Voluntary Compliance than to enforced Compliance; on the latter, the perception had no significant influence. This finding is inconsistent with the findings by Muehlbacher et al. (2011) Inasius (2019b), which show that trust in the tax authority was the strongest predictor on voluntary compliance. The weak influence of the perception of government's power might be related to SMEs' inability to comprehend the complexities of the tax system (Pope & Abdul-Jabbar, 2008) in terms of enforced tax compliance. Another possibility is that the respondents had already declared their unpaid tax through the tax amnesty program, because of which they did not fear sanctions. The lack of a statistically significant interaction effect between trust and power supports recent studies using the Slippery Slope Framework that found either no influence or only a small influence of this interaction on Table 3. Effects of Trust in the Government and Perception of Government's Power on Perception of Voluntary Tax Compliance (n = 410).

Independent variable
Attitude toward voluntary tax compliance Step 1 Step 2

Independent variable
Attitude toward enforced tax compliance Step 1 Step 2 voluntary or enforced tax compliance (Kogler et al.,2013(Kogler et al., , 2015Muehlbacher et al., 2011). However, the influence of the perception of government's power on the agreement with voluntary tax compliance was not as strong in this study as shown in earlier research (Kogler et al., 2013(Kogler et al., , 2015Muehlbacher et al., 2011), and the lack of any influence of attitude toward enforced compliance conflicted with the findings of earlier studies (Kogler et al., 2015;Muehlbacher et al., 2011). This difference might be explained by taxpayers' lack of any feelings of guilt after participating in tax amnesty. They were not concerned about sanctions from the tax authority because, based on amnesty law, taxpayers who participate in amnesty do not face audits for the previous years. Furthermore, for those who are highly compliant, audits not only are associated with the cost of taxation (such as time) but also create dissatisfaction, which might generate disappointment among people regarding the tax authority's policies. Low levels of taxpayers' satisfaction can, in turn, reduce the number of people who pay taxes because they feel the government should be making stronger efforts to increase the convenience of paying taxes. Perhaps, they expect greater governmental commitment to increasing services, rather than increasing the number of tax audits. Hence, a power-oriented approach is not always an effective method to increase tax compliance, as it might backfire and decrease tax compliance (Alm, 2018).
Accordingly, although increasing the government's power may force Indonesia's taxpayers to obey the tax laws, a policy that builds trust may be less costly and more effective (Muehlbacher et al., 2011;Slemrod, 2016). In addition, efficient and effective tax services and a simplified tax compliance process might increase taxpayers' trust in the government. Therefore, trustworthy taxpayers should be rewarded, and dishonest taxpayers should be prosecuted (Inasius, 2019a;Kirchler et al., 2008). The government's failure to enforce tax compliance on tax evaders might negatively affect compliant taxpayers. This reasoning is supported by the results of Allingham and Sandmo (1972), Kogler et al. (2013), andPrinz et al. (2014) on how social inequality in the taxation system facilitates tax evasion.
The current study's results indicate that age does not affect perceptions of enforced tax compliance, which generally support the findings of Kogler et al. (2015) and Muehlbacher et al. (2011). However, as highlighted by Kirchler et al. (2008), earlier studies' findings on the effects of age are inconsistent. Several studies found that age did not influence enforced compliance (Minor, 1978;Song & Yarbrough, 1978), whereas others reported significantly contradictory results (Kirchler, 1999;Vogel, 1974). Regarding other demographic characteristics, the differences between men and women and those among people with different educational background may be considered in the formulation of voluntary and enforced compliance policies.
Because age is positively related to voluntary compliance, auditing younger taxpayers would be a targeted method to prevent tax noncompliance. This conclusion supports the findings of Kirchler (1999), Muehlbacher et al. (2011), andVogel (1974). In contrast, for enforced compliance, auditing less educated taxpayers is a targeted method to detect tax evaders. This result contradicts the finding by Muehlbacher et al. (2011) that more educated people were less compliant. This difference may be explained by the factors that influence taxpayers' attitudes after participating in tax amnesty that are specific to developing countries, such as Indonesia. In such countries, factors such as weak legal institutions and weak administrative capabilities (Slemrod, 2016) may have more influence on educated taxpayers. Although more accurate knowledge on taxation may increase prudence regarding taxation-related obligations, educational attainment may be positively related to accurate knowledge about taxes, cause enforced compliance, and increase the belief that taxes are enforced. If more and better knowledge on taxes increases positive attitudes toward the taxation system, educational attainment will be positively related to tax compliance (Eriksen & Fallan, 1996).
According to earlier studies, tax amnesty has conflicting results; in some cases, amnesty has succeeded and, in others, it has failed (Bose & Jetter, 2012). Accordingly, a better understanding of how to encourage taxpayers to participate in tax amnesty can be used to increase tax revenues and tax compliance levels (Alm & Beck, 1993;Dunn et al., 2016;Torgler & Schaltegger, 2005). A tax amnesty might not be more beneficial than more stringent tax auditing. However, by better understanding taxpayers' behavior following a tax amnesty, tax authorities may be able to increase the collection of tax revenue and enhance subsequent compliance.
The findings of this study have the following implications. One possible implication of is that the study findings can help policymakers and taxation authorities formulate taxation policies and better predict taxpayers' behaviors after implementing a tax amnesty. The social implications of the study findings require more details on how different retail SMEs would react in an amnesty situation based on their demographic characteristics.
In summary, this study provides evidence that policies of enforced compliance are less effective than policies of voluntary compliance. However, to enhance voluntary compliance, steps must be taken to build trust. Hence, the maintenance of voluntary tax compliance seems to be more effective and less expensive, although it requires a synergistic environment in which the government and taxpayers can interact in mutually beneficial ways. The method to create such a healthy environment is to treat taxpayers fairly and to prosecute dishonest taxpayers. Accordingly, trust building is the main aspect of voluntary compliance policies, and auditing is a supportive policy that should focus on low-compliance taxpayers in developing countries, such as Indonesia.
Earlier studies on the Slippery Slope Framework have focused on developed countries (Kirchler et al., 2008;Kogler et al., 2013Kogler et al., , 2015Lisi, 2012;Muehlbacher et al., 2011;Prinz et al., 2014). For developing countries, Inasius (2019b) and Batrancea et al. (2019) formalized the Slippery Slope Framework of tax compliance in Indonesia. The results support empirical research and show that trust is an important factor in voluntary compliance. However, that research was conducted before the implementation of tax amnesty.
By examining the framework in a developing country following tax amnesty implementation, this study extends the literature by demonstrating that the influence of trust found in developed countries applies to developing countries, as well. Furthermore, supporting the literature on tax compliance in developed countries, this study demonstrated that a power-based approach is not the best method to achieve tax compliance. Tax compliance is likely to increase through government's efforts to increase and maintain voluntary compliance by emphasizing certainty and fairness in tax laws, transparency in governmental decisions (Slemrod, 2016), and a service-oriented tax administration that relies on trust in a friendly and synergic relationship with taxpayers. Therefore, this study demonstrates that the important role played by trust in developed countries is mirrored in developing economies following the implementation of a tax amnesty.
This study has several limitations. First, the sample comprised SME retailers alone; it completely excluded nonretailer taxpayers. Therefore, the results are probably not generalizable to taxpayers other than SME retailers. Second, the data collection conducted through face-to-face interviews might be biased, particularly because the information sought is sensitive and related to tax evasion (Bowling, 2005;Szolnoki & Hoffmann, 2013). Hence, future studies should investigate other possible determinants and use other data collection methods to extend the current study's findings. One potentially fruitful approach may be to compare the tax compliance of nonretailers and retailers and improving data collection through a mail survey.

Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding
The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was supported by Bina Nusantara University (Grant 014/ VR.RT/III/2018).