Assessing Taxpayer Behavior Commitment in Mattering Tax Compliance: The Moderating of Tax Ethics

: This study aims to determine the effect of affective commitment, normative commitment, and ongoing commitment to tax compliance with tax ethics as a moderating variable. This research is quantitative research using primary data. This study's population is individual taxpayers registered at KPP Pratama Kosambi, Tangerang City, Indonesia. The sampling technique used convenience sampling, which obtained a sample of 100 respondents—collecting data using a questionnaire/questionnaire method. The measurement scale used is the Likert scale. The data analysis technique used in this research is a quantitative data analysis using statistical methods. The statistical method used is Partial Least Square (PLS). The results of this study indicate that normative commitment has a significant effect on tax compliance, while affective and ongoing commitments do not affect tax compliance. Tax ethics cannot moderate the relationship between affective, normative, and sustainable commitment to tax compliance.


LITERATURE REVIEW
Tax compliance requires active participation from taxpayers in fulfilling their tax obligations, which necessitates a high level of taxpayer compliance, which involves accurately meeting tax obligations (Susyanti et al., 2021). The concept of tax compliance can be understood in two ways: a simple definition focuses on the extent to which taxpayers meet their tax obligations as prescribed, while a comprehensive definition emphasizes responsible taxpayer behavior as citizens, going beyond mere fear of sanctions (Adhikara et al., 2022). According to Drews & Van den Bergh (2016), individual behavior, including tax compliance, is influenced by various factors, including predisposing factors (such as knowledge, attitudes, and beliefs), enabling factors (such as the physical environment and available resources), and reinforcing factors (such as laws, regulations, and supervision). Commitment plays a crucial role in understanding taxpayer behavior, with affective commitment driven by a strong desire to take action, normative commitment rooted in the sense of obligation and adherence to societal norms, and continuance commitment influenced by the perceived costs of non-compliance (Cahyonowati et al., 2020). Tax ethics, which encompasses beliefs and values regarding honesty and responsibility in taxation matters, also shapes taxpayer behavior (Yuniarta & Purnamawati, 2020). Understanding the relationship between tax compliance, commitment, and individual behavior is essential for policymakers and tax authorities in promoting voluntary tax compliance and effective tax administration (Saad, 2014). The study proposes several hypotheses regarding the influence of different types of commitment on tax compliance. Bornman & Wessels (2017) assert that affective commitment, characterized by emotional attachment to organizational values, is believed to positively influence tax compliance due to the belief that paying taxes is a moral obligation. Normative commitment, emphasizing the sense of obligation and responsibility to adhere to tax regulations, is expected to positively impact tax compliance. Then, continuance commitment, driven by the perceived costs of non-compliance, is hypothesized to motivate individuals to fulfill their tax obligations. Furthermore, tax ethics is proposed to moderate the effects of affective, normative, and continuance commitment on tax compliance. Higher ethical standards reinforce the positive influence of commitment on tax compliance. Based on the literature review provided by previous experts and researchers, the hypotheses formulated for this study are as follows: H1: Affective commitment has an impact on tax compliance. H2: Normative commitment influences tax compliance. H3: Continuance commitment affects tax compliance. H4: Tax ethics moderates the relationship between affective commitment and tax compliance. H5: Tax ethics moderates the relationship between normative commitment and tax compliance. H6: Tax ethics moderates the relationship between continuance commitment and tax compliance. These hypotheses are developed based on the existing theoretical frameworks and empirical evidence in tax compliance. This study aims to investigate the relationships between different types of commitment (affective, normative, and continuance) and tax compliance and explore the moderating role of tax ethics in these relationships. The study intends to provide insights into the factors influencing taxpayer compliance behavior and the potential moderating effect of ethical considerations through empirical analysis and statistical testing.

METHODOLOGY
The subjects of this study are individual taxpayers registered in KPP Pratama Kosambi, Tangerang City, Indonesia. They are influenced by independent variables: affective commitment, normative commitment, and continuance commitment. The population and sample of this study consist of all individual taxpayers registered in KPP Pratama Kosambi. The sampling technique used in this research is non-probability sampling, a precise convenience sampling method. As a result, a sample of 100 respondents who met the predetermined criteria was obtained for this study. Slovin's formula was used to calculate the sample size in this research. The outline of the research instrument can be seen in the following table.  The potential penalties and fines associated with noncompliance greatly influence my decision to adhere to regulations.  I am aware of the financial consequences of noncompliance, which motivates me to comply with the rules.  I consider the potential legal and reputational risks of non-compliance and prefer to avoid them.  The existence of strict penalties and fines serves as a deterrent for me to engage in non-compliant behavior.  I recognize that there are limited opportunities or alternatives to non-compliance, and therefore choose to comply with the established regulations.
 The potential penalties and fines for non-compliance are a strong deterrent, influencing me to comply with the regulations.  I understand non-compliance's financial and legal risks and strive to avoid them.  Due to the limited opportunities for non-compliance and lack of viable alternatives, I adhere to the established rules.  I consider the potential negative consequences, such as reputational damage, that may result from noncompliance.  The absence of viable alternatives or loopholes leaves me with no choice but to comply with the regulations.

Data Description
Based on the available data, this study has a higher representation of male respondents. Approximately 55% of the respondents are male, while 45% are female. It suggests a gender bias in selecting respondents, with a more significant proportion of male participants. In terms of age, the majority of the respondents fall within the 21-30 age range. They account for 74% of the total respondents. The remaining respondents are distributed among different age groups, including 31-40 (15%), 41-50 (9%), and above 50 (2%). Therefore, the study primarily focuses on individuals in the 21-30 age group. Regarding occupation, a significant portion of the respondents (74%) are employed in the private sector. A few respondents also identify as entrepreneurs (5%) or selfemployed (4%). It is worth noting that the specific occupation of some respondents should have been mentioned in the study (14%).

Outer Model
Construct validity can be assessed through various parameters such as the Loading Factor, Average Variance Extracted (AVE), and Communality. A construct is considered valid if the Loading Factor score is more significant than 0.5, AVE is greater than 0.5, and Communality is more significant than 0.5. The correlation between indicators and their respective constructs is displayed in Table 2. After conducting the validity test, the model's goodness-of-fit was evaluated based on the output model fit. The model obtained a value of 0.089 for SRMR, which is below the threshold of 0.100. Thus, the model is considered to be a good fit. The NFI value was also 0.638, above 0.5, but still far from the ideal value of 1, indicating that the model is relatively weak (see Table 3). The discriminant validity of a measurement model with reflective indicators is assessed based on the cross-loading of measurements with constructs. Suppose the correlation between a construct and its measuring items is higher than the correlation with other constructs. In that case, the latent construct predicts the measurements in its block better than those in other blocks (see Table 4). The cross-loading values presented in Table 4 provide evidence of solid discriminant validity. It is demonstrated by the higher correlation values observed between the indicators and their respective constructs compared to the correlation values between the indicators and other constructs. These findings suggest that the measuring items effectively capture the unique variance of their intended constructs, indicating a clear distinction between the latent constructs in the measurement model.

Inner Model
The structural model testing, or inner model, is conducted to examine the relationships between constructs, assess their significance, and determine the R-square values of the research model. The structural model is evaluated by using R-square for the dependent constructs under examination and the significance of the path coefficient results. When assessing the structural model with Partial Least Squares (PLS), the R-square values for each latent dependent variable can be examined. The estimation results of the R-square values obtained using PLS are presented in Table 5 below. Upon examining the information provided in Table 4.8, it becomes apparent that the derived R-square value of 0.634 suggests a substantial and robust relationship. The R-square value of 0.634 for tax compliance indicates that approximately 63.4% of the variation in the dependent variable can be elucidated by the independent variable considered in this study. This finding underscores the independent variable's significant impact and predictive power on tax compliance. However, it is essential to note that the remaining 36.6% of the variance in tax compliance is attributed to other variables not encompassed in the current research investigation. These unaccounted factors may include external influences, contextual aspects, or additional variables not incorporated in the study design.
Hypothesis testing examines the structural model (inner model), which involves analyzing the path coefficients indicating the parameter coefficients and their corresponding T-statistics. The significance of the estimated parameters provides valuable information for understanding the relationships between research variables. The basis for hypothesis testing relies on the values obtained from the output of path coefficients, as presented in the following Figure 1 and Table 6.   Table 45, the hypotheses related to affective commitment, normative commitment, continuance commitment, and the moderating effect of tax ethics were examined concerning tax compliance.
For affective commitment, the result of 1.863 with a T-statistic value below 1.96 led to the rejection of the first hypothesis (H1), indicating that affective commitment does not significantly influence tax compliance. It suggests that lower levels of affective commitment correspond to lower levels of tax compliance. However, these findings differ from a study by Bornman & Wessels (2017), which reported a weak positive correlation between affective commitment and tax compliance.
Regarding normative commitment, the result of 2.551 with a T-statistic value above 1.96 supports the second hypothesis (H2) acceptance, indicating a significant influence of normative commitment on tax compliance. It implies that higher levels of normative commitment contribute to increased tax compliance, reflecting an alignment between individuals' sense of obligation and responsibility to adhere to tax regulations. However, this finding contradicts the weak correlation reported by Bornman & Wessels (2017) and emphasizes the role of different implementation approaches for normative commitment indicators in tax compliance studies.
Concerning continuance commitment, the result of 1.732 with a T-statistic value below 1.96 leads to rejecting the third hypothesis (H3), suggesting that continuance commitment does not significantly affect tax compliance. Lower levels of continuance commitment correspond to lower levels of tax compliance, indicating that individuals' awareness of potential losses for noncompliance, such as penalties or punishments, does not strongly influence their tax payment behavior. However, this differs from the findings of Bornman & Wessels (2017), who reported a moderate correlation between continuance commitment and tax compliance.
The moderating effects of tax ethics were also examined. The results indicate that tax ethics do not moderate the relationship between affective commitment and tax compliance (H4), normative commitment and tax compliance (H5), or continuance commitment and tax compliance (H6). These findings align with a study by Yuhertina et al. (2016) and highlight the limited moderating role of tax ethics in influencing tax compliance. However, these results contradict the findings of Oktaviani & Saifudin (2019), emphasizing the need to consider different implementations of ethical indicators in tax compliance studies.
In summary, this study found that affective commitment does not significantly influence tax compliance, while normative commitment significantly affects tax compliance. Continuance commitment does not significantly impact tax compliance, and tax ethics do not moderate the relationship between commitment types and tax compliance. These results highlight the complex nature of tax compliance and the importance of considering various factors when studying taxpayer behavior.

CONCLUSION
In conclusion, this study aimed to examine the relationships between commitment types (affective, normative, and continuance) and tax compliance and the moderating effects of tax ethics. The findings shed light on the intricate dynamics within tax compliance.
Firstly, the results indicated that affective commitment does not significantly influence tax compliance. It suggests that individuals' emotional attachment or connection to tax obligations does not substantially shape their compliance behavior. On the other hand, normative commitment emerged as a significant factor influencing tax compliance. Higher levels of normative commitment, driven by a sense of obligation and responsibility, were associated with greater adherence to tax regulations. However, the study revealed that continuance commitment only significantly impacts tax compliance. The awareness of potential losses, such as penalties or punishments, did not strongly influence individuals' compliance with tax obligations. The fear of negative consequences alone may not be a strong motivator for ensuring tax compliance.
Furthermore, the analysis of the moderating effects of tax ethics demonstrated that tax ethics did not significantly moderate the relationships between commitment types and tax compliance. The influence of affective, normative, and continuance commitment on tax compliance remains consistent regardless of ethical considerations. Overall, this study highlights the multifaceted nature of tax compliance. While normative commitment is crucial in fostering tax compliance, affective and continuance commitments may have little direct effect. Moreover, the findings suggest that tax ethics may not substantially affect the relationship between commitment types and tax compliance. These findings contribute to understanding taxpayer behavior and have implications for policymakers and tax authorities in designing effective strategies to promote tax compliance. By recognizing the significance of normative commitment and considering additional factors beyond emotions and fear of consequences, policymakers can develop targeted interventions and initiatives to enhance taxpayer compliance. This research has several limitations, including: 1) The study only sampled individual taxpayers registered at the Kosambi Pratama Tax Office, limiting the generalizability of the findings to a specific population. The results may be different from the broader taxpayer population. 2) In the data collection process, the information respondents provide through questionnaires may only sometimes reflect their accurate opinions. It could be due to differences in thinking, understanding, and varying assumptions among respondents.
Additionally, factors such as respondent honesty in completing the questionnaires can impact the accuracy of the data collected. 3) There may be limitations in previous research on the topic, resulting in a need for more available references to support and enhance the current study. The limited references could impact the depth and breadth of the analysis and discussion.