THE VOLUNTARY DISCLOSURE DILEMMA: UNRAVELING THE COMPLIANCE-EVASION CAUSALITY IN TAX ADMINISTRATION

This research investigates the causality between taxpayer compliance and tax evasion behaviors, specifically within the context of participants in the Voluntary Disclosure Program (PPS) registered at the Small Tax Office of West Pontianak. The study delineates its population as taxpayers who, prior to their engagement in the PPS, had outstanding tax liabilities on income derived from business or employment activities. Utilizing the documentation method, secondary data were solicited from pertinent governmental bodies to facilitate the research. A linear regression model was employed to analyze the relationship between the variables under consideration. The findings underscore the impact of pre-PPS tax evasion activities on subsequent enhancements in taxpayer compliance, as evidenced by ransom payments. The study contributes to governmental authorities by offering valuable information regarding the patterns of tax evasion behavior among PPS participants, thereby informing policy and enforcement strategies.


INTRODUCTION
Tax avoidance constitutes a legal strategy within the ambit of tax planning, characterized by the lawful structuring of fiscal affairs to minimize income tax liabilities.This strategy exploits extant legal loopholes, enabling taxpayers to circumvent adverse legal repercussions, such as penalties or sanctions, arising from tax avoidance maneuvers (Barli, 2018;Oktavia et al., 2021).Although tax avoidance and tax evasion are both aimed at diminishing tax liabilities, they diverge fundamentally in legality.Tax evasion involves the illicit reduction or negation of tax obligations through unauthorized means, distinguishing it markedly from tax avoidance (Barli, 2018;Purba et al., 2022).Saputri and Kamil (2021) delineate various tax evasion tactics, including the failure to report accurate assets and income, the misalignment of tax payments with statutory requirements, and the omission of periodic or annual tax returns.Additionally, Purba et al. (2022) observe that tax evasion can extend to the strategic placement of assets in jurisdictions with favorable tax regimes-often referred to as tax havens-or countries that offer reduced tax rates or tax exemptions.These evasion practices compromise taxpayer compliance, potentially precipitating significant revenue losses for the state (Anam et al., 2018;Monica & Andi, 2019;Riyadi et al., 2021), thereby underscoring the critical distinction between legal tax avoidance measures and illicit tax evasion actions.
Analyzing the behavior and characteristics of individual taxpayers reveals multiple determinants influencing their propensity towards tax evasion (Ekaputra et al., 2022;Nathalie & Setiawan, 2024).These factors encompass perceptions of fairness, experiences of discrimination, and attitudes towards the tax system (Sasmita & Kimsen, 2023), a predilection for material wealth or the belief among taxpayers that tax payments are futile and financially detrimental (Umaimah, 2021;Zainuddin et al., 2021), as well as the taxpayers' income levels, which reflect their economic capabilities (Randiansyah et al., 2021).Furthermore, the inclination to evade taxes is also shaped by the manner in which tax regulations are applied and executed by governmental authorities, including the quality of public services, the efficacy of the implemented tax system, and the enforcement of penalties for non-compliance (Kamil, 2021).In response to these challenges and in a bid to enhance compliance with tax reporting obligations, the Indonesian government has instituted a tax amnesty policy (Inasius et al., 2020).This policy aims to encourage the declaration of previously unreported net assets, serving as a proxy for an increase in the taxpayers' economic status or income (Ispriyarso, 2019).The amnesty provides relief from administrative and criminal penalties for undeclared income, contingent upon the payment of a defined amount (referred to as "ransom") based on the taxes applicable to the newly disclosed net assets (Kusuma & Dewi, 2018;Nugraha & Setiawan, 2018).This approach seeks not only to rectify past non-compliance but also to foster a more transparent and cooperative relationship between taxpayers and the tax authorities (Hadistiyah & Putra, 2022;Wulan et al., 2023).Kurniawan et al. (2019) articulate that the tax amnesty initiative is designed to achieve both immediate and protracted objectives.In the near term, it is anticipated to bolster the fiscal year's tax revenue through the collection of "ransom" payments on newly disclosed net assets (Darma et al., 2022;Mardi, 2019).Over a more extended period, the initiative seeks to cultivate a culture of enhanced compliance among taxpayers with regard to their reporting duties.This strategic shift is aimed at diminishing the prevalence of tax evasion, broadening the tax base, and fostering economic growth via the reallocation of assets (Murweni, 2018;Pravasanti, 2018).The underlying rationale for the tax amnesty, as aligned with the overarching objective of securing increased tax revenue, hinges on the principle that elevated levels of taxpayer compliance will directly contribute to higher tax revenue collections, predicated on lawful taxpayer behaviors and the avoidance of tax evasion (Riyadi et al., 2021).Nonetheless, research conducted by Purba et al. (2022) casts doubt on the efficacy of the tax amnesty program in mitigating tax evasion within Indonesia, revealing a persistent inclination towards such practices even amidst the policy's enactment.This inclination is exemplified by a notable surge in the allocation of funds to offshore banking institutions, exceeding 137 million USD, which suggests that the amnesty's implementation has not necessarily translated into improved taxpayer compliance (Hermawan et al., 2020;Permana, 2020).The persistence of tax evasion behaviors subsequent to the tax amnesty policy underscores the complexity of ensuring compliance through policy measures alone (Kurniawan et al., 2019), highlighting the necessity for comprehensive strategies that address the underlying factors contributing to evasion (Sayidah & Assagaf, 2019).
Extant literature on tax evasion predominantly explores the determinants prompting taxpayers to engage in such practices, with notable contributions from Kamil (2021), Randiansyah et al. (2021), Sasmita and Kimsen (2023), Umaimah (2021).In parallel, scholarly inquiry into the tax amnesty policy's implementation has largely focused on evaluating its impact on taxpayer compliance levels (Mardi, 2019).Beyond compliance metrics, other investigations have assessed the tax amnesty policy's influence on the efficacy of tax revenue collection (Suratno et al., 2020).Regarding the objective of mitigating tax evasion through tax amnesty, limited research, such as the study by Purba et al. (2022), has examined the correlation between tax amnesty initiatives and shifts in taxpayer conduct, particularly in terms of increased overseas fund allocations.
Diverging from the aforementioned scholarly endeavors, this study aims to investigate the nexus between tax evasion behaviors and enhanced taxpayer compliance, specifically through the mechanism of ransom payments under a voluntary disclosure program.This research seeks to contribute to the academic discourse by elucidating the potential for ransom payments to not only signal but also catalyze a transformation in taxpayer compliance, thereby offering new insights into the dynamics between tax evasion practices and complianceenhancing strategies.The novelty of this study lies in the variables selected for analysis and the methodology employed to assess the interrelations among these variables.Specifically, the research endeavors to ascertain the influence of tax evasion behaviors on the magnitude of ransom payments made by participants in the Voluntary Disclosure Program, utilizing these payments as a proxy for heightened taxpayer compliance.The metric for evaluating an increase in compliance is operationalized through the ransom amounts levied on previously undisclosed assets (Riyadi et al., 2021), whereas the gauge for tax evasion intensity is based on the undeclared tax liabilities associated with business or employment income that taxpayers have failed to remit (Saputri & Kamil, 2021).
This exploration suggests a complex, possibly non-linear, relationship between the original intents underpinning the tax amnesty policy and the subsequent shifts in taxpayer attitudes and behaviors post-implementation.The study seeks to offer a novel perspective on the dynamics between the behavioral predispositions of taxpayers enrolled in the Voluntary Disclosure Program towards tax evasion, and the extent to which ransom payments reflect and potentially alter these tendencies.The critical inquiry revolves around whether the tax amnesty policy's execution effectively mirrors and modifies the propensities of taxpayers inclined towards evasion.Addressing this query necessitates an empirical examination of the correlation between enhanced compliance among Voluntary Disclosure Program participants and their evasion activities.The findings are anticipated to serve as a valuable reference for governmental bodies, specifically the Directorate General of Taxes, in formulating targeted oversight strategies for taxpayers predisposed to evasion, thereby informing policy adjustments and enforcement frameworks.

LITERATURE REVIEW
Hagger (2019) explains that in the Theory of Reasoned Action (TRA), a person's intention is a motivational foundation that has a major influence in determining a person's behavior.These intentions build individual attitudes based on the results of evaluating the negative and positive impacts of individual attitudes and subjective norms in society that view how individuals should behave towards their environment.This theory is used to predict how individuals will behave towards a problem or condition based on interests that are influenced by beliefs about the results of past events and views from other individuals on the same problem or condition.In general, taxpayers have a tendency to pay the lowest tax possible, even if possible they will try to avoid it (Margaretha et al., 2023).Umaimah (2021) explains that there are two factors that can affect taxpayer compliance in disclosing assets owned, namely internal factors related to the taxpayer's lack of understanding of the benefits or usefulness of fulfilling tax obligations and external factors in the form of information from outside parties with negative connotations related to the management and implementation of tax policies.Another study conducted by Mujiyati et al. (2022) concluded that taxpayers who participate in the tax amnesty program have a higher tendency to commit tax evasion and are more prone to tax evasion than taxpayers who do not participate in the tax amnesty program where the higher the level of disclosure of net assets and payment of ransom, the higher the level of tax evasion or evasion committed by taxpayers.The description above explains that there is a relationship between taxpayer attitudes and behavior and tax evasion or avoidance.
The implementation of the tax amnesty program has been implemented several times in Indonesia, namely the Tax Amnesty Period I policy in 1964, the Tax Amnesty Period II policy in 1984, the Tax Amnesty Period III policy in 2007, the Tax Amnesty Period IV policy in 2009, the Tax Amnesty Period V policy in 2015 and the Tax Amnesty Period IV policy in 2016.In 2021, the government through Law Number 7 of 2021 concerning Harmonization of Tax Regulations provides an opportunity for Individual and Corporate Taxpayers participating in tax amnesty to re-disclose assets that have not been reported or not reported at the time of participating in the tax amnesty program through policy I Voluntary Disclosure Program and Individual Taxpayers other than tax amnesty participants to disclose net assets still owned on December 31, 2020 which were obtained from January 1, 2016 to December 31, 2020 and have not been reported in the Individual Annual Tax Return for the 2020 Fiscal Year through policy II Voluntary Disclosure Program.The purpose of enacting this policy is to increase taxpayer voluntary compliance (Mahmud & Mooduto, 2023;Ningtyas & Aisyaturrahmi, 2022).To encourage the successful implementation of the program, the government offers compensation or benefits for Voluntary Disclosure Program participants.By participating in the Voluntary Disclosure Program, Taxpayers participating in PPS Policy I will not be subject to administrative sanctions of an increase of 200% of undisclosed assets and Taxpayers participating in PPS Policy II will receive benefits including no tax audit on tax obligations for Fiscal Years 2016 to 2020 and data and information from the disclosure of net assets cannot be used as the basis for criminal investigation, investigation and/or prosecution.
According to Irawan and Raras ( 2021), the Voluntary Disclosure Program can be referred to as tax amnesty volume II because it has the same substance, namely the provision of tax amnesty for net assets that have not been disclosed in the Tax Return.Amnesty.This PAS-Final policy aims to provide an opportunity for taxpayers participating in tax amnesty to make improvements to asset disclosure reporting if there are assets that have not been fully disclosed when participating in the tax amnesty program and taxpayers not participating in tax amnesty to disclose assets that have not been reported in the Annual Tax Return (SPT) (Farhan & Rosdiana, 2023).By participating in the PAS-Final program, taxpayers can avoid the imposition of an increase sanction of 200% of the value of assets that have not been or less disclosed for tax amnesty participants and 2% per month for a maximum of 24 months starting from the discovery of data and / or information on additional income until the issuance of an Underpaid Tax Audit Letter (SKPKB) for taxpayers not participating in tax amnesty.
From the description above, the question arises, the provision of opportunities for taxpayers to correct or re-report assets that have not been disclosed on the first opportunity to report the assets when participating in the tax amnesty program through the implementation of the PAS-Final program in 2017 and the Voluntary Disclosure Program in 2022 can fully encourage voluntary compliance from taxpayers or even have the opposite effect?According to the results of research from Ispriyarso (2019) the existence of legal uncertainty in the application of tax amnesty sanctions creates a tendency for taxpayers not to pay taxes in advance and prefer to wait for other tax amnesty policies in the future because the ransom payment from the tax amnesty program is considered cheaper.The results of the study show that on the other hand, the existence of tax amnesty can even trigger the disobedient attitude of taxpayers.
Based on the explanation above, it is known that taxpayers in determining their attitudes and behavior in carrying out tax obligations are based on individual reasons or intentions to minimize tax payments, which is indicated by the behavioral tendency to commit tax evasion (Sasmita & Kimsen, 2023;Umaimah, 2021;Zainuddin et al., 2021;Ispriyarso 2019).This tendency of tax evasion can be seen from the behavior of taxpayers who choose not to pay taxes in advance and prefer to wait for future tax amnesty policies in order to pay taxes that are considered cheaper through ransom payments (Ispriyarso 2019).The delay in payment behavior has an impact on the amount of ransom payments when participating in the tax amnesty program, where the delay in large tax payments will lead to greater disclosure of net assets when participating in the tax amnesty program so that the ransom which is considered an increase in taxpayer compliance will be high.Based on the description above, the first hypothesis (H1) can be formulated in the form of an alternative hypothesis as follows: the increase in taxpayer compliance through the payment of ransom is influenced by the level of tax evasion.

METHOD
This research is anchored in the positivist paradigm, which serves as the conceptual foundation for elucidating the phenomena and realities inherent to the topic under investigation.Within the framework of positivism, it is posited that empirical facts form the exclusive basis for all scientific assertions, with social reality perceived as objective (Wekke, 2019).Guided by this paradigmatic stance, the researcher employs a quantitative methodology to quantitatively assess and interpret the correlation between taxpayer compliance and tax evasion (Ambarwati et al., 2021;Qadri et al., 2023;Qadri & Darmawan, 2021).The investigative process is operationalized through a case study approach, leveraging quantitative methods to examine the behaviors and responses of taxpayers enrolled in the Voluntary Disclosure Program at KPP Pratama Pontianak Barat.
The analytical focus of this investigation encompasses individuals and entities enrolled in two distinct cohorts of the Voluntary Disclosure Program: Policy I targets Individual and Corporate Taxpayers who participated in the Tax Amnesty program yet failed to fully declare their net assets up to December 31, 2015.Simultaneously, Policy II pertains to those who acquired undisclosed net assets between January 1, 2016, and December 31, 2020, and omitted these from their 2020 Tax Return (SPT), thereby potentially engaging in tax evasion through the non-remittance of taxes due on earnings from business or employment activities.Data for this study were procured via the secondary data documentation technique, involving formal requests for relevant data from designated agencies, specifically KPP Pratama Pontianak Barat, in accordance with the data requisition protocols established by the Directorate General of Taxes.This process entailed the submission of a comprehensive data request alongside an application for research authorization through the online platform www.eriset.pajak.go.id.The application process required the provision of several documents, including an endorsement or introductory letter from the affiliated academic institution, a detailed research proposal, and a formally stamped declaration committing to the dissemination of the research findings to the Directorate General of Taxes.
The total population comprising taxpayers enrolled in Policies I and II of the Voluntary Disclosure Program at KPP Pratama Pontianak Barat numbers 773, as delineated in Table 1.For sample selection, this study employs the Purposive Sampling technique, a methodological approach whereby specific individuals or instances are deliberately chosen to yield critical insights unattainable from alternative sources.This selection process, as articulated by Firmansyah & Dede (2022), incorporates cases or participants into the research sample with the intent of generating findings that align with the study's initial aims and accurately reflect the characteristics of the broader population under investigation.One of the practices of tax evasion is by not paying the tax burden in accordance with the provisions of the law (Saputri & Kamil, 2021) The sample criteria specify that taxpayers eligible for inclusion in the study are those engaged in the Voluntary Disclosure Program, possessing income from business or employment activities, yet who have either failed to report or remit tax obligations on said income for the tax years 2018 to 2020, or for any tax year prior to the commencement of the Voluntary Disclosure Program.Of the 773 taxpayers who participated in the Voluntary Disclosure Program, it was found that 731 of them did not indicate having committed tax evasion (no unpaid tax data).Thus, taxpayers who meet the criteria to become research samples are 42 taxpayers.
This research delineates two primary variables: taxpayer compliance, serving as the dependent variable, and tax evasion, positioned as the independent variable.The operationalization and measurement of these variables are comprehensively detailed in Table 2.The Taxpayer Compliance Variable (TPC) measures the increase in tax revenue through ransom payments (Riyadi et al., 2021), while the Tax Evasion Variable (TEV) measures unpaid tax obligations (Saputri & Kamil, 2021).The study applies a single linear regression model to evaluate the impact of the independent variable (tax evasion) on the dependent variable (taxpayer compliance), with the model's formulation grounded in the operationalization outcomes of the variables under examination.lnTPC = β0 + β1*lnTEV + ε In the process of conducting quantitative data analysis, the researcher begins by compiling a dataset that includes information on the total tax liabilities unpaid by taxpayers during the tax years 2018 to 2020, alongside data pertaining to the ransom payments made in relation to the disclosure of net assets within the Voluntary Disclosure Program.Once the dataset is prepared, the subsequent step involves selecting an appropriate regression technique, for which a single linear regression method is employed.Following the selection of this method, the researcher undertakes a comprehensive regression analysis.This analysis encompasses the execution of diagnostic tests to assess measurement errors, verify the regression model's specification, and conduct tests for classical assumptions, which include examinations of normality and heteroscedasticity.These procedural steps are critical to ascertain that the chosen model adheres to the classical assumptions, thereby ensuring that the resulting coefficients are BLUE (Best Linear Unbiased Estimate), indicative of the most reliable and unbiased estimates achievable within the linear regression framework.
The measurement error test is conducted with the objective of verifying the precision with which the variables have been quantified.To assess the accuracy of these measurements, an exhaustive descriptive analysis is performed for each variable to examine the distribution characteristics, utilizing the skewness and kurtosis coefficients as indicators (Cain et al., 2017).Moreover, an analysis of outlier data is undertaken through the application of the predicted Cook's distance value for variables exhibiting signs of non-normal distribution, specifically identified by a skewness coefficient divergent from zero and a kurtosis coefficient exceeding three (Smiti, 2020).Variable data manifesting a Cook's distance value surpassing one are subsequently excluded from the research sample to maintain the integrity of the analysis.Upon ensuring the appropriate measurement of the sample and variables, the analysis proceeds to the regression specification test and the classical assumption test.The regression specification test employs the scatter plot method to ascertain the linearity of the regression model and verify that the relationship between variables aligns with the formulated hypothesis.This step is crucial in validating that the regression model is correctly specified and that it adherently reflects the theoretical relationship posited between the dependent and independent variables.
The subsequent phase in the analysis encompasses the regression specification test and the testing of classical assumptions.The regression specification test is executed via the scatter plot technique to assess the linearity of the regression model and to confirm its conformity with the predefined hypothesis, ensuring the model's appropriateness for the data and theoretical expectations (Nguyen et al., 2020).In addition to the regression specification, a series of classical assumption tests are conducted to validate the underlying assumptions of the regression model .This includes the normality test (Ruxton et al., 2015), implemented through the Shapiro-Wilk method, to verify the normal distribution of the dataset.Concurrently, a heteroscedasticity test (Romeo et al., 2023), utilizing the Breusch-Pagan method, is employed to ascertain the constancy of error variance or predictive error across the dataset, aiming to establish homoscedasticity.
Given the employment of a single linear regression model in this investigation, the test for multicollinearity was deemed unnecessary (Kim, 2019).Multicollinearity testing is typically relevant in models involving multiple independent variables, where high correlations among predictors may distort the reliability of the regression coefficients.Furthermore, the analysis did not encompass an autocorrelation test, predicated on the rationale that the dataset is cross-sectional (Pötscher & Preinerstorfer, 2018).Cross-sectional data, representing observations at a single point in time, inherently minimizes the concerns of autocorrelation typically associated with time-series data, where the independence of observations across time intervals is a critical assumption.

RESULT AND DISCUSSION
Table 3 presents a summary of the descriptive statistics of each variable used in the study.The sample of the research conducted was 42 taxpayers who were taxpayers participating in the Voluntary Disclosure Program who were indicated to have committed tax evasion as indicated by the existence of unpaid tax data on income received or earned during the 2018 tax year period up to the 2020 tax year.However, after analyzing the accuracy of data measurement through the measurement error test using the skewness and kurtosis coefficient indicators in the initial descriptive analysis, it is known that the skewness coefficient value of the two research variables is greater than 0 (zero) and the kurtosis coefficient value of the taxpayer compliance variable is greater than 3 (three) which indicates the presence of outliers which causes the selected sample data not to be normally distributed.
To overcome the abnormal data distribution, the author changes the unit of measure for each variable data, which was originally in the form of the rupiah amount of the ransom value for the taxpayer compliance variable and the unpaid tax value for the tax evasion variable, to the natural logarithm (Ln) unit of measure.Furthermore, to be able to determine the presence of outliers, a prediction of the cook's distance value on each variable is carried out using the STATA statistical data processing software.From the prediction of the cook's distance value, it is known that there are 2 (two) variable data that are indicated as outliers (cook's distance value greater than 1 (one)) so that the data is removed from the research sample and the remaining samples to be used as research data and statistical analysis using STATA software are 40 samples.The exclusion of outlier data from the research sample is based on the consideration that after the outlier data is removed, the kurtosis coefficient value of the taxpayer compliance variable which was previously greater than 3 (three) changes to less than 3 (three) and the skewness coefficient value of both variables is close to 0 (zero) which indicates that the data is normally distributed.
Based on Table 3, it can be concluded that of the 40 samples of taxpayers studied, the taxpayer compliance variable proxied by the amount of ransom paid when participating in the voluntary disclosure program (lnTPC) has the smallest (minimum) natural logarithm value of 14.853 or IDR 2,823,365 and the largest (maximum) value of 21.326 or IDR 1,827,141,886.The average (mean) amount of ransom paid by the sampled taxpayers has a natural logarithm (Ln) value of 18.055 or IDR 69,411,978 and a median value of 17.830 or IDR 55,433,796.The natural logarithm (Ln) value of the deviation of 1.425 which is smaller than the average value indicates that the taxpayer compliance variable has low data variation.
In the tax evasion variable (lnTEV) proxied by the total amount of tax that has not been paid or paid by taxpayers for income received or earned in the period 2018 to 2020 tax year, it is known that of the 40 sample taxpayers studied, the smallest (minimum) natural logarithm value (Ln) is 14.200 or IDR 1,469,643 and the largest (maximum) value is 20.172 or IDR 576,738,779.The average (mean) of the total amount of tax evasion from the sampled taxpayers has a natural logarithm (Ln) value of 16.582 or IDR 15,907,480 and a median value of 16.712 amounting to IDR 18,110,456.The natural logarithm (Ln) value of the deviation of 1.567 which is smaller than the average value indicates that the tax evasion variable has low data variation.

. Regression Specification Test Scatter Plot
Figure 1 above is the result of the regression specification test using the Scatter Plot method which illustrates the linear and positive relationship between the taxpayer compliance variable as the dependent variable and the tax evasion variable as the independent variable.The linear and positive relationship is depicted by a straight line on the graph which has a positive slope, indicating that the relationship between the independent variable and the dependent variable is directly proportional or linear.The relationship between variables based on the graph above is in accordance with the hypothesis compiled where the graph illustrates a directly proportional or linear relationship between tax evasion and increased taxpayer compliance where taxpayers participating in the PPS with a high level of tax evasion will pay a high ransom which indicates an increase in taxpayer compliance in making tax payments when participating in the Voluntary Disclosure Program.
Based on the results of the Normality test using the Shapiro-Wilk method in accordance with Table 4 above, it is known that the probability value (Prob (z)) of the taxpayer compliance variable is 0.469 and the tax evasion variable is 0.227.The Prob (z) value of each variable is greater than 0.050 so that it can be ascertained that the data used in the study are normally distributed and the normality assumption has been fulfilled.With the fulfillment of the assumption of normality, further statistical analysis can be carried out on the data that is the research sample.
In addition, the table above shows the results of the Heteroscedasticity test using the Breusch-Pagan method where the probability value (Prob (Chi-sq)) is 0.145.From the Prob (Chi-sq) value, it can be ascertained that the research data is homoskedasticity (not heteroskedasticity) because the Prob (Chi-sq) value is greater than 0.050.Based on the results of the heteroscedasticity test, it is known that the error variance or prediction error of the research data is constant throughout the data range.Based on the regression test results Table 4 and Table 5, a single linear regression equation model is obtained as follows.lnTPC = 6,977 + 0,668*lnTEV + ε Based on the single linear regression equation above, it is known that the constant value of 6.977 is positive, which indicates that if the independent variable of tax evasion is assumed to be 0 (zero), then the dependent variable of taxpayer compliance with the indicator of ransom payment is the natural logarithm value (Ln) 6.977.The tax evasion variable (lnTEV) has a coefficient of 0.668 which indicates that a 1.000% increase in the amount of tax evasion (lnTEV) will increase the amount of ransom payments as an indicator of increased taxpayer compliance (lnTPC) when participating in the Voluntary Disclosure Program by 0.668%.0.000 Table 5 shows the coefficient of determination (adjusted R 2 ) of 0.527 which indicates that the effect of the independent variable of tax evasion on the dependent variable of taxpayer compliance is 52.700% while the remaining 47.300% is influenced by other variables not included in the study.H1 states that the increase in taxpayer compliance through ransom payments is influenced by the level of tax evasion.Judging from the significance value of the independent variable tax evasion of 0.000 or less than the value of α = 0.050, it can be concluded that the tax evasion variable has a significant positive effect on the taxpayer compliance variable.From the significance and nature of the relationship, it can be concluded that H1 is accepted or rejects H0 and it can be concluded that the higher the tax evasion committed by taxpayers participating in the Voluntary Disclosure Program, the higher the increase in taxpayer compliance when participating in the Voluntary Disclosure Program as indicated by the greater the amount of ransom paid when participating in the Voluntary Disclosure Program.
In general, taxpayers have a tendency to pay the lowest possible tax or even if possible will try to avoid it (Margaretha et al., 2023).To overcome the tendency of tax avoidance that leads to tax evasion behavior that can reduce potential tax revenue for the state (Anam et al., 2018;Monica & Andi, 2019;Riyadi et al., 2021), the government through the Directorate General of Taxes has implemented a tax amnesty policy with the hope of increasing the level of taxpayer compliance through increased tax revenue from ransom deposits on net assets disclosed when participating in the tax amnesty program.However, on the other hand, the increase in taxpayer compliance through ransom payments may indicate how the tax evasion behavior of taxpayers who participate in the tax amnesty program.
The results showed that there was a significant effect of every 1.000% increase in the amount of tax evasion on the increase in the amount of ransom of 0.668%.From the results of this study, it can be seen that taxpayers with a higher amount of ransom payments or experiencing a high increase in compliance when participating in the PPS program tend to have a high level or amount of evasion.This is in line with the results of research conducted by Mujiyati et al. ( 2022) which states that taxpayers who participate in the tax amnesty program have a higher tendency to commit tax evasion and are more prone to tax evasion compared to taxpayers who do not participate in the tax amnesty program where the higher the level of net asset disclosure and ransom payments, the higher the level of tax avoidance or evasion committed by taxpayers.
Based on the literature review that has been conducted, there are several factors that become reasons or causes that can lead to tax evasion behavior (reasoned action).These factors include perceptions of fairness, discrimination and the tax system (Sasmita & Kimsen, 2023), the nature of love of money or the assumption of taxpayers that tax payment is something that is not useful and causes losses (Umaimah, 2021;Zainuddin et al., 2021) and the existence of legal uncertainty in the application of tax amnesty sanctions (Ispriyarso, 2019).These factors form a tendency for taxpayer behavior not to pay taxes first and prefer to wait for another tax amnesty policy in the future because the ransom payment from the tax amnesty program is considered cheaper (Ispriyarso, 2019).From the explanation above, it can be concluded that the existence of a tax amnesty program implemented by the government can actually trigger tax evasion behavior as indicated by the behavior of taxpayers who do not report and pay taxes that are their obligations directly or when they are owed taxes and choose to report them when participating in the tax amnesty program.The tendency to report income or increase net assets only when participating in the tax amnesty program is certainly not in line with the original purpose of implementing the program, namely to reduce the level of tax evasion (Kurniawan et al., 2019).This condition is also supported by other research conducted by Purba et al. (2022) which states that tax evasion practices actually tend to continue during the implementation period of the tax amnesty program.
The tendency of tax evasion committed by taxpayers participating in the tax amnesty program is in line with theory of reasoned action.At the time of participating in the Voluntary Disclosure Program, taxpayers with a high level of tax evasion will pay ransom in a larger amount compared to taxpayers who do not commit tax evasion or taxpayers with a lower level or amount of tax evasion.The existence of a larger ransom payment indicates that the taxpayer's income measured by the addition of net assets that were not or have not been reported partially or wholly before participating in the Voluntary Disclosure Program, is not reported correctly or in accordance with the actual conditions so that when participating in the tax amnesty program will cause a large amount of ransom.The decision to report additional net assets when participating in a Voluntary Disclosure Program is a reasoned action (Hagger, 2019).The main reason for this preference is the advantage of a tax burden that is considered lower through the ransom payment (Ispriyarso, 2019).

CONCLUSION
The findings of this research elucidate that tax evasion exerts a positive impact on taxpayer compliance within the context of the Voluntary Disclosure Program.This positive correlation is manifested in the observation that the higher the level of tax evasion engaged in by taxpayers prior to their participation in the Voluntary Disclosure Program, the more substantial the ransom payments derived from the disclosure of net assets.Such payments serve as indicators of enhanced taxpayer compliance, consequent to the implementation of the Voluntary Disclosure Program.Nonetheless, this correlation also suggests that taxpayers who make significant ransom contributions likely exhibited a pronounced propensity towards tax evasion before their engagement with the Voluntary Disclosure Program.Previous studies have previously investigated the nexus between tax amnesty policies and shifts in taxpayer behaviors and attitudes, particularly using the metric of increased offshore fund allocations by taxpayers as a proxy for the propensity towards tax evasion.The current study extends this inquiry by examining the relationship between taxpayers' predispositions towards tax evasion, quantified by unpaid tax liabilities, and their compliance levels within the Voluntary Disclosure Program.Specifically, the aim is to ascertain the extent to which prior tax evasion activities influence subsequent compliance improvements, as evidenced by the volume of ransom payments associated with the disclosure of net assets.Through this lens, the research seeks to contribute to a nuanced understanding of the dynamics between pre-program tax evasion behaviors and compliance enhancements facilitated by participation in the Voluntary Disclosure Program.
This research furnishes the government, specifically the Directorate General of Taxes, with insights into the tax evasion tendencies among taxpayers enrolled in the Voluntary Disclosure Program.The findings are intended to inform the Directorate General of Taxes, underscoring the need for enhanced oversight of taxpayers within this program who exhibit propensities towards tax evasion, with the ultimate goal of maximizing state revenue.Despite its contributions, this study acknowledges limitations in the scope of its research variables and the operationalization of these variables.The analysis reveals that the independent variables considered herein account for only 52.700% of the variance in the dependent variable.Moreover, the operational measure of tax evasion-based on the aggregate amount of undeclared taxes by participants of the Voluntary Disclosure Program-is confined to the tax years spanning 2018 to 2020.Future research is encouraged to expand upon the present study by incorporating additional relevant variables and extending the timeframe for variable measurement indicators.Such extensions would potentially offer a more comprehensive understanding of taxpayer behavior over an extended period, thereby enhancing the predictive power and applicability of the research findings to policy formulation and enforcement strategies aimed at curbing tax evasion and improving taxpayer compliance.
After the implementation of the tax amnesty program and before the enactment of the Voluntary Disclosure Program, a similar program has been implemented in the form of the Voluntary Asset Disclosure program using the Final Rate (PAS-Final) through Minister of Finance Regulation Number 165/PMK.03/2017 on the Second Amendment to Minister of Finance Regulation 118/PMK.03/2016 on the Implementation of Law Number 11 of 2016 on Tax

Table 1 . Purposive Sampling Method Criteria Sample Taxpayers
participating in the Voluntary Disclosure Program 773 Taxpayers participating in the Voluntary Disclosure Program are not indicated to have committed tax evasion ( )