The Effect of Transfer Pricing, Tunneling Incentive, Thin Capitalization, and Capital Intensity against Tax Avoidance

This study aims to investigate the influence of transfer pricing, tunneling incentives, thin capitalization, and capital intensity on tax avoidance practices in mining companies listed on the Indonesia Stock Exchange (IDX) during the period of 2018-2022. The context of this study lies in the complexity of global taxation and increasing concerns over tax avoidance practices that have prompted the need for a deeper understanding of the factors influencing corporate behavior in this regard. The research method employed involves the use of statistical regression to measure the relationship between these variables. The analysis results show that tunneling incentives have a significantly negative effect on tax avoidance practices, while thin capitalization and capital intensity significantly influence tax avoidance practices positively. Meanwhile, transfer pricing does not have a significant effect on tax avoidance practices. These findings suggest that tax avoidance practices tend to be carried out by companies employing thin capital structures and having significant business activities to invest their assets in fixed assets. In this context, the study provides valuable insights for regulators and practitioners in designing policies aimed at reducing tax avoidance practices. By understanding the factors influencing the company's decision in this regard, more effective prevention and control measures can be implemented. In conclusion, this study emphasizes the importance of paying attention to the interaction between transfer pricing, tunneling incentives, thin capitalization, and capital intensity in the context of corporate tax avoidance to develop a holistic


Introduction
Taxes are one of the most important sources of state revenue in carrying out the functions of government and development of a country.However, in its implementation, companies often carry out various strategies to reduce the tax burden they have to pay.One strategy that is often used is tax avoidance, which refers to companies' efforts to avoid paying as much tax as possible by taking advantage of loopholes in applicable tax regulations.In the Tax Justice Network report entitled "The State of Tax Justice 2020: Tax Justice in the time of Covid-19," it was revealed that around 4.78 billion US dollars or equivalent to IDR 67.6 trillion came from corporate tax avoidance practices in Indonesia (Marino & Nicolodi, 2021).Taxes play an important role in a country's financial framework, and in Indonesia in particular, taxes have a significant important role as the largest source of revenue and the highest contributor to the State Budget (APBN) (Suryantari & Mimba, 2022) The researchers' reasons for choosing the mining sector are based on the data presented in Table .1 Tax ratio is the ratio between the tax revenue of a particular sector or sectors with the total national tax revenue in a country.In the 2018-2022 period, it can be seen that the national tax ratio, which is the percentage of tax revenue from all economic sectors, has decreased significantly from 11.5% in 2018 to 8.3% in 2020, before rising again to 10.4% in 2022.
Meanwhile, the mining sector showed more dramatic fluctuations in the tax ratio compared to other sectors, falling dramatically from 4.95% in 2018 to only 1.22% in 2020, before increasing sharply to 7.1% in 2022.In addition, if the average calculation is carried out in the last five years, the average contribution of mining is the lowest at 4%, compared to manufacturing (10%), construction (5%), and transportation (5%).Therefore, the mining sector is interesting to investigate further to understand the factors that influence fluctuations in its tax contribution and its impact on overall national tax revenue.
This research itself is also based on the phenomenon of transfer pricing which is a form of tax avoidance practices that occur in Indonesia, especially in mining companies.In 2019, Global Witness disclosed the results of an investigation into PT Adaro Energy being involved in transfer pricing practices.PT Adaro has a subsidiary in Singapore called Coaltrade Services International.Their analysis concluded that through Coaltrade, PT Adaro from 2009-2017 had paid a lower tax burden than it should, with a difference of USD 125 million.The next phenomenon is mimicked from indications of tax avoidance in companies with the largest revenues, namely PT Adaro Energy Indonesia Tbk (ADRO), PT Golden Energy Mines Tbk (GMES), PT Indo Tambangraya Megah Tbk (ITMG).From the comparison of the value of Earnings Before Tax (EBT) and Book Tax Differences (BTD) during the 2018-2022 period, there are interesting trends to analyze.First, it can be seen that the value of NRE has experienced significant fluctuations from year to year, with a fairly sharp increase from 2020 to 2021 and 2022.This may reflect changes in the company's operational performance or external factors such as changes in regulations or market conditions.Meanwhile, BTD values also show significant variations from year to year.The presence of negative BTD values in all years indicates that fiscal profit (used for taxation) tends to be greater than accounting profit.This could indicate that the company may be taking advantage of various tax incentives or tax reduction mechanisms available.Also, note that the value of BTD is getting more negative year by year, which may indicate an increase in tax deductions or tax avoidance activities.Nonetheless, the sharp decline in the value of BTD from 2021 to 2022, from -0.025 to -0.076, could be a concern as it may indicate a change in corporate tax management strategy or a change in tax regulation.
The study of various phenomena and research motivations above is a strong basis for urgency related to the topic of tax avoidance.Analysis of the factors influencing the taxation behavior of mining companies, especially related to tax avoidance practices, is essential to understand this phenomenon more deeply and explore the urgency of research in this regard.To further understand why these mining companies, engage in tax avoidance practices, it is necessary to analyze the factors that influence their tax behavior.Some of the relevant factors in this context are transfer pricing, tunneling incentive, thin capitalization, and capital intensity.
Transfer pricing is a company's practice of determining the price of buying and selling between affiliates that can affect the amount of profit reported in various jurisdictions (Suandy, 2016).Guidelines related to transfer pricing are regulated in the Regulation of the Minister of Finance of the Republic of Indonesia Number 22/PMK.03/2020concerning Procedures for Implementing Transfer Price Agreements (Advance Pricing Agreement).This regulation replaces Minister of Finance Regulation Number 7/PMK.03/2015.In addition, article 18 paragraph (4) of Law No. 7 of 2021 concerning Harmonization of Tax Regulations (HPP Law) also regulates Transfer Pricing practices.The results of research by Lutfia & Pratomo (2018) stated that transfer pricing had a positive impact on tax avoidance efforts, while research by Suryantari & Mimba (2022) states that transfer pricing has no effect on tax avoidance efforts.
The concept of tunneling incentive itself refers to the incentive of major shareholders or related parties to shift profits from companies that go public (companies whose shares are listed on the Indonesia Stock Exchange) to related companies that are not listed (OpraVita, 2023).This finding is in accordance with the results of research conducted by Asmaradana (2021) and Lestari & Solikhah (2016), which stated that tunneling incentives have a positive impact on tax avoidance efforts.However, research conducted by the study indicates that (Nurulita &Yulianto, 2023) tunneling incentives have a negative influence on tax avoidance efforts.
The next factor is thin capitalization which refers to a situation when a company funds itself with a high level of debt compared to the capital it has, otherwise known as highly leveraged One tax avoidance strategy is to use intergroup loans from countries with lower tax rates.The results of research by Olivia & Dwimulyani (2019) & Waluyo & Doktoralina (2019) show that thin capitalization has a positive impact on tax avoidance efforts.However, research by Nainggolan & Hutabarat (2022) shows that thin capitalization does not have a significant impact on tax avoidance efforts.ISSN 2721-0960 (Print), ISSN 2721-0847 (online)  Copyright © 2024, Journal La Sociale, Under the license CC BY-SA 4.0 The last factor is capital intensity which shows how much the company invests in the form of fixed assets.Capital intensity is considered to affect the amount of tax paid by the company.Several previous studies such as those conducted by Suciarti et al. (2020) and Darsani & Sukharta (2021) show that capital intensity has a positive impact on tax avoidance efforts, where the higher the capital intensity in a company, the greater the tax avoidance carried out by the company.However, different studies, such as those conducted by Marlinda et al. (2020) and Pattiasina et al. (2019), show that capital intensity actually has no influence on tax avoidance efforts.
The results of previous research studies on these factors in fact still show inconsistencies or research gaps.With this background, this study aims to analyze the effect of transfer pricing, tunneling incentive, thin capitalization, and capital intensity on tax avoidance practices in mining companies listed on the IDX.The results of this study are expected to provide further insight into what factors influence companies' decisions in conducting tax avoidance, as well as the implications of this practice on operational efficiency and financial position of companies in this sector.Thus, this research will not only make theoretical contributions in the accounting and taxation literature, but also have practical implications for companies and regulators in developing better policies related to tax avoidance practices in mining companies in Indonesia.Based on this presentation, it aroused the interest of researchers to conduct research with the title: "The Effect of Transfer Pricing, Tunneling Incentive, Thin Capitalization, and Capital Intensity on Tax Avoidance in Mining Companies on the Indonesia Stock Exchange (IDX) for the 2018-2022 Period".

Methods
In this study, the type of research used is associative research with a quantitative approach.This approach was chosen because the data to be used in this study is quantitative, meaning that the research carried out is research that emphasizes its analysis on numerical data (numbers) which are then processed and analyzed for conclusions.This research uses secondary data, which only takes data that has been processed and presented by other parties and downloaded from the Indonesia Stock Exchange website and from each company website included in the research sample.The population used in this study is entities from mining companies listed on the Indonesia Stock Exchange in 2018-2022.The sampling technique in this study uses purposive sampling, which is a sampling technique of data sources with certain criteria.Data analysis methods in this study include descriptive statistical analysis, classical assumption test, multiple linear regression test and hypothesis test.This test used Microsoft Excel application and Software Statistical Package for the Social Sciences (SPSS) version 27.The results of the normality test table analysis showed that the data were not normally distributed, with a significance value (p-value) of less than 0.05 (0.000c).Furthermore, one effort that can be made to address this is by eliminating outlier data.This can be done by identifying and removing extreme values that may cause abnormality in the distribution of data.

Normality test
Below are the results of outlier data observations.Based on Figures 2, 3, 4, 5, and 6, it is concluded that there are a total of 9 (nine) extreme data points, which are excluded from the research sample.Subsequently, a normality test was conducted again with the following results.The results of the multicollinearity test showed that the tolerance values for all independent variables (X1 RPT, X2 TI, X3 MAD, X4 CAPIN) were above the recommended limit of 0.10, with a range between 0.701 to 0.893.In addition, the Variance Inflation Factor (VIF) value is also below the threshold value of 10, with a range between 1.120 to 1.427.Based on this criterion, it can be concluded that there is no multicollinearity among independent variables.That is, there are no serious problems in the linear relationship between the independent variables in the regression model used.Thus, the results of the analysis show that the assumption of multicollinearity is fulfilled in the regression model performed.The results of the heterokedasticity test above show points that tend to be scattered towards the left on the Y axis with a significance value of less than 0.05.This indicates that the pattern is negative.Then, one attempt to overcome this is with the help of the Glejser test to make the data spread towards the right which is positive.The basis for making decisions of the Glejser test is that if the significance value > 0.05, then the data do not indicate symptoms of heterokedasticity.This test can be done by progressing the independent variable to the absolute value of the residual on the independent variable.The following are the results of heterokedasticity testing after the Glejser additional test.pattern or a wide pattern that signifies heteroscedasticity.Figure 8 shows that error variability is constant across the range of predicted values, which shows that the homoscedasticity assumption is satisfied in the regression model.Therefore, it can be concluded that there is no statistical evidence to suggest the existence of heteroscedasticity in the regression model used.The results of the autocorrelation test with the Durbin-Watson method that has been transformed through lag (Cochrane-Orcutt) show that the statistical value of Durbin-Watson obtained after transformation is 1.954 (dw).While the calculated significance t value for N amounts to 46 with 4 independent variables is 1.7201 (dU) and the value of 4-(1.3448)dL is 2.6552.Thus, the results of this test state that dU<dw<(4-dL) or 1.7201<1.954<2.6552,so it can be concluded that there is no autocorrelation either positive or negative.The results of multiple regression analysis in Table 7 show that the regression model has the following equation: Tax Avoidance (BTD) = -0.007+ 0.004RPT -0.030TI + 0.003MAD + 0.039CAPIN + ε

Multiple Linear Regression Analysis Test
The constant intercept value is -0.007.This result can be understood that if the magnitude of all independent variables is 0, then the magnitude of Book Tax Differences (BTD) is -0.007.The variable X1 RPT has a regression coefficient of 0.004, this value indicates a positive direction.Thus, an increase of one unit in the variable X1 RPT will cause an increase of 0.004 in the dependent variable tax avoidance (Y BTD), assuming the other variables remain constant.The variable X2 TI has a regression coefficient of -0.030.This coefficient is negative, indicating that an increase of one unit in the variable X2 TI will lead to a decrease of 0.030 units in the dependent variable tax avoidance (Y BTD) assuming the other variable remains constant.
The variable X3 MAD has a regression coefficient of 0.003.This suggests that the variable X3 MAD has a positive influence.Thus, an increase of one unit in the variable X3 MAD will cause an increase of 0.003 in the dependent variable tax avoidance (Y BTD), assuming the other variable remains constant.The variable X4 CAPIN has a regression coefficient of 0.039.This coefficient is positive, indicating that an increase of one unit in the variable X4 CAPIN will lead to an increase of 0.039 units in the dependent variable tax avoidance (Y BTD) assuming the other variables remain constant.

Partial Effect Test (T Test)
Table 7 records the results of partial influence tests relevant to this study.The following are the results of a partial effect analysis for each independent variable on the dependent variable tax avoidance (BTD).In variable X1 because the p value (0.197) is greater than the established significance level (0.05), it is statistically insufficient evidence to be able to conclude that transfer pricing has a significant positive effect on tax avoidance (BTD).Ha 1 was rejected.In the variable X2 because the p value (0.000) is smaller than the established significance level (0.05), it is statistically sufficient evidence to conclude that tunneling incentive has a significant effect on tax avoidance (BTD), but the direction of the regression coefficient shows a negative influence while what is expected on the development of the hypothesis is a positive direction so that Ha. 2 is rejected.
In variable X3 because the p value (0.032) is smaller than the established significance level (0.05), it is statistically sufficient evidence to conclude that thin capitalization has a significant positive effect on tax avoidance (BTD).Ha 3 accepted.In variable X4 because the p value (0.000) is smaller than the established level of significance (0.05), the researcher has enough evidence to conclude that capital intensity has a significant positive effect on tax avoidance (BTD).Ha 4 accepted.Based on the results of the analysis, it can be concluded that MAD and CAPIN variables have a significant positive effect on the tax avoidance (BTD) dependent variable, while TI variables have a significant negative influence.In addition, the RPT variable has a positive but not significant influence.independent variable has a significant influence on the dependent variable in this regression model.Thus, it can be concluded that the regression model used in this analysis, all independent variables namely transfer pricing (X1), tunneling incentive (X2), thin capitalization (X3), and capital intensity (X4) simultaneously have a significant effect on tax avoidance (Y).Based on the table the value of R Square is 0.449, which means that about 44.9% of the variability in Y BTD can be explained by X1 RPT, X2 TI, X3 MAD, and X4 CAPIN.From these results, although the model as a whole was able to explain some of the variability in the dependent variable, there was still about 55.1% variability that could not be explained by this model.This can be an area for model improvement or refinement.

The Effect of Transfer Pricing on Tax Avoidance
The first hypothesis proposed regarding transfer pricing having a positive effect on tax avoidance was rejected.This is because hypothesis testing is not statistically significant even though the direction of the regression coefficient has shown a positive direction.There are several factors that lead to this rejection.Descriptive statistical analysis shows that the mean of variable transfer pricing (RPT) is only 0.13501, or only 13.5% of the company's total receivables in the sample.This indicates that companies in the classified sample must have a relationship with overseas operations, showing transactions that tend to be relatively small in terms of receivables compared to domestic transactions.Based on the analysis, transfer pricing, which involves pricing transactions between related parties, may not directly impact tax avoidance for several reasons.First, when viewed from accounting standards, PSAK 7 is one of the references related to the disclosure of related parties.This standard does not provide detailed regulation of transfer pricing practices in disclosing transactions with related parties.Second, regulations such as Minister of Finance Regulation No. 213/PMK.03/2016instruct taxpayers to keep certain documents and information related to transactions with related parties to minimize transfer pricing practices.Based on the results of statistical studies of sample data and related regulations, these reasons are used as a justification for researchers related to why the proposed hypothesis was rejected.The findings of this study are in line with research by Suryantari & Mimba (2022) Irawan et al. (2020) & Sutanto & Lasar (2023) which states that transfer pricing has no effect on tax avoidance.

The Effect of Tunneling Incentive on Tax Avoidance
The findings of this study reject the hypothesis proposed that tunneling incentives have a positive effect on tax avoidance.Conversely, in this study it is stated that tunneling incentives have a significant negative effect on tax avoidance.So, this research is not in line with agency theory where one of the agency problems that arises is triggered by differences in interests between controlling and minority shares (Nurulita & Yulianto, 2023).
The rejection of the hypothesis proposed in this study can be indicated through observation of statistical data related to tunneling incentive (IT) against tax avoidance (BTD).In mean, IT ISSN 2721-0960 (Print), ISSN 2721-0847 (online) Copyright © 2024, Journal La Sociale, Under the license CC BY-SA 4.0 sample data companies amounted to 58.24% which indicates that the largest proportion of shareholding (controller) in the sample data dominates compared to minority shares.The maximum IT value owned by PTRO occurred in 2022 with a percentage of 89.79% controlling shares compared to a minority, the BTD figure actually showed -0.0009 which indicates low tax avoidance practices.Unlike the case with BIPI which has a minimum IT value of 23.57%, the BTD value owned in the current year is -0.002 which actually indicates greater aggressiveness in tax avoidance than PTRO.This explains that there is a finding why TI has a negative coefficient because the greater the proportion of controlling shares actually decreases the value of BTD or indicates low tax avoidance practices.
According to Nurulita & Yulianto (2023) findings, tunneling incentives negatively affect tax avoidance because controlling shareholders can take advantage of their privileges to transfer company assets and profits to other companies they control in countries with lower tax rates.This can reduce the taxes owed by the company and increase the profits of controlling shareholders.However, the higher the proportion of controlling shares, the smaller tax avoidance practices will be because minority shareholders will pay more attention to the interests of the company and limit tax avoidance practices that harm the company.This finding is also in line with Susanti & Firmansyah (2018) & Rahmawati & Mulyani (2020) which states that tunneling incentives negatively affect tax avoidance.

The Effect of Thin Capitalization on Tax Avoidance
The hypothesis proposed states that thin capitalization (MAD) has a positive effect on tax avoidance.The results of the hypothesis test show statistical significance, which confirms that there is a strong relationship between MAD and tax avoidance practices.Descriptive statistical support states that the average MAD of companies in the sample is fairly high, namely the average debt compared to Safe Harbor Debt Amount (SHDA) is 0.7080 or 70.8%.These finding states that the greater the thin capitalization, the greater the tax avoidance practice with the higher the value of BTD.This is because the lower the ratio of own capital to debt, the more likely the company is to take action to reduce tax payments.Managers tend to take advantage of the high interest burden of debt to reduce taxable profits, thereby increasing tax avoidance practices.
This finding is consistent with previous studies, such as research by Olivia & Dwimulyani (2019) & Waluyo & Doktoralina (2019), which also found that thin capitalization has a positive impact on tax avoidance efforts.These findings confirm that companies with thin capitalization tend to have high interest expenses, which can ultimately reduce taxable profits and increase tax avoidance practices.Thus, the acceptance of the hypothesis that thin capitalization has a significant positive effect on tax avoidance shows the importance of paying attention to the company's capital structure in the context of tax policy and financial arrangements.

The Effect of Capital Intensity on Tax Avoidance
Capital intensity, which refers to the ratio between fixed assets and total assets of a company, is key in this dynamic.Simply put, the greater the capital intensity ratio (CAPIN), the greater the potential for tax avoidance.From the perspective of agency theory, the relationship between capital intensity and tax avoidance can be well explained.This theory assumes the existence of a conflict of interest between the owner and management of the company.Management may tend to optimize the company's profits, including by reducing the tax burden.Companies that have high fixed assets have a relatively low tax burden compared to companies that have low fixed assets (Marlinda et al., 2020).
Research findings support this hypothesis that capital intensity has a significant positive effect on tax avoidance.Descriptive statistical analysis shows that companies with higher CAPIN ratios tend to have greater BTD values.For example, TBS Energi Utama Tbk. has the lowest ISSN 2721-0960 (Print), ISSN 2721-0847 (online) Copyright © 2024, Journal La Sociale, Under the license CC BY-SA 4.0 CAPIN ratio of 0.02825 and BTD value of -0.0015, while Soechi Lines Tbk.has the highest CAPIN ratio with 0.87575 and BTD value of 0.003.The mean CAPIN is 0.31137, with a standard deviation of 0.27027.This indicates that the greater the CAPIN ratio, the greater the potential for tax avoidance.
In addition, support from previous studies by Dharma & Novaria (2017) & Sari & Indrawan (2022) reinforces this conclusion.Their research shows that capital intensity has a positive impact on tax avoidance efforts.For example, large investments in fixed assets can result in depreciation costs that can reduce the tax burden.These findings are in line with previous research and provide a strong empirical foundation for this hypothesis.Thus, it can be concluded that the greater the capital intensity of a company, the greater the company's efforts to avoid taxes, in accordance with agency theory and existing empirical findings.

Conclusion
Transfer pricing has no significant effect on tax avoidance.Therefore, the hypothesis that there is a positive influence between transfer pricing and tax avoidance is rejected.Tunneling incentives turned out to have a significant negative influence on tax avoidance.That is, the lower the tunneling incentive, the higher the likelihood of the company evading taxes.The hypothesis of both studies was rejected.Thin capitalization has a significant positive influence on tax avoidance.This confirms that companies tend to use thin capital structures to maximize tax avoidance.The third hypothesis of the study was declared accepted.Capital intensity also has a significant positive influence on tax avoidance.This suggests that the greater the level of capital intensity, the higher the likelihood of a company using tax avoidance strategies.The fourth hypothesis of the study was declared accepted.

Figure 8 .
Figure 8. Heteroscedasticity Test of Scatterplot Glejser-Test Source: SPSS 27 output, processed by researchers (2024)The results of the heteroscedasticity test using the scatterplot technique showed that there was no indication of heteroscedasticity symptoms in the regression model analyzed after the Glejser transform test.The scatterplot between the dependent variable and the independent variable shows a uniform point scattering pattern and does not show a specific pattern such as a cone

Table 1 .
National and Mining Tax Ratio . ISSN 2721-0960 (Print), ISSN 2721-0847 (online) Copyright © 2024, Journal La Sociale, Under the license CC BY-SA 4.0 another taxpayer's capital.Meanwhile, paragraph 4 of Article 18 regulates taxpayers who directly or indirectly control one or more other taxpayers they control jointly.
Figure 1.Comparison of BTD & EBT of Three Largest Mining Companies in IndonesiaThis practice causes the potential state revenue to be reduced by USD 14 million annually.This figure should be used by the state to support national development and public interest.The phenomenon related to the issue of transfer pricing has now also been explained in Law Number 7 of 2021 concerning Harmonization of Tax Regulations, especially in Article 18 paragraphs (3) and (4).This relationship refers to situations where a taxpayer generates at least 25% of another taxpayer's income or has direct or indirect ownership of as much as 25% of EBT BTD ISSN 2721-0960 (Print), ISSN 2721-0847 (online) Copyright © 2024, Journal La Sociale, Under the license CC BY-SA 4.0

Table 3 .
Normality Test Results After Outlier EliminationThe analysis results of Table3for normality test after outliers indicate that the data are normally distributed, with a significance value (p-value) greater than 0.05 (0.200c).

Table 5 .
Autocorrelation Test Results

Table 7 .
Multiple Linear Analysis Test Results

Table 8 .
Simultaneous Effect Test Results (Test F) Source: SPSS 27 output, processed by researchers(2024)The results of the simultaneous influence test(Test F)show that the regression model as a whole has a significant influence on the dependent variable Y1.An F-statistic value of 191.243 with a significance value (p-value) of less than 0.05 (sig p = 0.000), indicates that at least one ISSN 2721-0960 (Print), ISSN 2721-0847 (online) Copyright © 2024, Journal La Sociale, Under the license CC BY-SA 4.0

Table 9 .
Test Results of Coefficient of Determination (R²)