Tax avoidance in the Indonesian manufacturing industry

This study aims to determine tax avoidance in the Indonesian manufacturing industry. The control variable is the industrial sub-sector. In this study tax avoidance is proxied as book tax differences. This study focuses on manufacturing companies listed on the Indonesia Stock Exchange (IDX) for the 2015-2018 period. Using the STATA 13 regression data panel, this study shows that capital intensity has no significant effect on tax avoidance and inventory intensity has no significant effect on tax avoidance. However, this study found that industry auditors' specialization had a positive effect on tax avoidance. This study contributes to enhancing empirical evidence of audit quality, proxied by industry auditors' specialization can provide assurance of managers' tax avoidance activities did not violate the fairness principle, thus increase tax avoidance activities.


Introduction
Taxes are the largest source of state revenue. Where 82.5% of Indonesia's total revenue comes from tax revenues. Tax revenue in 2019 amounted to 1,545.3 trillion from the 2019 target of 1,577.6 trillion. This tax revenue figure grew 1.7% from the realization in 2018 (Kementerian Keuangan, 2018). Tax revenue is used to carry out and support national development operational activities such as infrastructure, education, health, social affairs, and others for the welfare of the Indonesian people. In this case, the role of taxes has a big influence on the source of state revenue, the government seeks to increase revenue in the tax sector. The current implementation of taxation is carried out by companies (taxpayers). It is seen that there are still many companies that do tax avoidance by taking advantage of loopholes in tax regulations. Usually, the company does so by minimizing the tax burden which will reduce the company's profit segment. The difference in the interests of the company makes the Director General of Taxes have to make updated tax regulations in order to minimize manufacturing companies taking advantage of existing regulatory loopholes.
The obstacle faced in maximizing current tax revenue is tax avoidance activities by companies (taxpayers) (Swingly & Sukartha, 2015). The Directorate General of Taxes found tax avoidance by PT. Coca Cola Indonesia. The Director General of Taxes conducted audits for cases in 2002,2003,2004. The Directorate General of Taxes (DGT) found a large increase in costs that year. The value of this large expense causes less taxable income so that the tax payment is also small. This expense is used for Coca Cola advertisements from 2002-2006 with a total of Rp. 566.84 billion. As a result of experiencing a reduction in taxable income, the taxable entrepreneur (PKP) of PT. Coca Cola Indonesia must deposit Rp. 603.48 billion to the state, but according to the calculations of PT. Coca Cola Indonesia only deposits Rp. 492.94 billion to the state. This figure has a difference of Rp.49.24 billion that is not paid up. The Directorate General of Taxes suspects that there is transfer pricing conducted by PT. Coca Cola Indonesia (kompas.com). Transfer pricing is a principle of fairness in transactions between taxpayers and parties who have a special relationship (Dharma & Noviari, 2017). The Toyota Motor Manufacturing company also carried out transfer pricing in 2005. The Directorate General of Taxes was suspicious and examined the Annual Tax Return (SPT). The company sends its manufactured goods abroad because the tax rate is lower than the tax rate applicable in Indonesia.
Tax avoidance is also experienced by Starbuck, the British tax official reports that Starbuck claimed massive losses during 2008-2010. It turned out that in reports to investors, he experienced a big profit. Starbucks can also legally move profits overseas. This is done, among others, by offshore licensing. This strategy deals with intellectual property rights. Every year Starbuck UK deposits its profits to Starbuck Netherlands (BBC News, 2012). Starbuck avoids paying its taxes in England to benefit the company by transferring royalties to Starbuck Holland. offshore licensing is a form of tax avoidance.
Tax avoidance is legal because it does not violate taxation provisions. Usually, companies avoid their taxes so that the profits obtained from company activities can be used to increase company finances. Factors that influence include tax avoidance, namely capital intensity, inventory intensity, and specialized industry auditors. The capital intensity in the company, namely fixed assets. In fixed assets, there is a depreciation expense. Depreciation expenses can reduce the tax burden owed by the company. So that companies that have a high level of fixed assets have a lower tax burden than companies that have low fixed assets. According to Dharma and Noviari (2017) and Irianto et al. (2017) that capital intensity has a positive effect on tax avoidance. Companies that have high fixed assets can minimize the tax burden payable from the depreciation expense of fixed assets each year. Meanwhile, according to Budianti and Curry (2018), capital intensity has a negative effect on tax avoidance and according to Puspita and Febrianti (2017) and Putra and Merkusiwati (2016), capital intensity has no effect on tax avoidance.
High inventory levels will create additional burdens for the company. Additional expenses include storage costs, administrative costs, labor costs, material costs, and sales costs. These costs will be a deduction from the tax burden that the company will pay. But according to Dwiyanti and Jati (2019) and Manurung (2019), inventory intensity has a positive effect on tax avoidance. Anindyka et al. (2018), inventory intensity has a negative effect on tax avoidance. However, inventory intensity has no significant effect according to Lestari (2015) because part of the investment is not the right way to do tax management.
Another factor influencing tax avoidance is the specialty of the auditor industry. This auditor industry specialization has to do with audit quality results. Public accounting firms that have specialist industry expertise auditors can assist audited companies (clients) in planning tax avoidance. According to Lee and Kao (2018), the specialty of industry auditors has an effect on tax avoidance because auditors can help companies avoid their taxes. According to Khairunisa et al. (2017), Setiyawati and Rohman (2015), Amaliyah and Rachmawati (2019), auditor specialization has a negative effect and Salehi et al. (2020) assess that auditor specialization has no effect on tax avoidance. An auditor will maintain his reputation and not cooperate with companies to evade tax. Based on this research, the formulation of the problem of this research is whether there is an effect of capital intensity, inventory intensity, and industry auditors' specialization on tax avoidance. While the purpose of this study was to determine Capital Intensity, Inventory Intensity, Auditors Industry Specialization Effect of Tax Avoidance.

Literature Review Agency Theory
Agency theory is a theory that occurs with a relationship between the party giving authority (the principal) and the party receiving the authority (agent), the principal is the company owner and the agent is the manager (Muzakki & Darsono, 2015). The agent is obliged to report and provide information to the principal because an agent has a better understanding of the company's condition than the principal. However, sometimes the authorized party (agent) does not report the actual situation experienced by the company. Managers as agents seek to fulfill the contractual relationship of shareholders as principals by choosing accounting methods or policies that take the interests of shareholders into account (Arieftiara & Yanthi, 2017). Agency theory in the context of this research is that managers must provide clear information about the condition of the company to shareholders so as not to cause information asymmetry. Where the corporate taxpayers calculate and report their taxes. In this collection system, companies can minimize their taxes by taking advantage of loopholes in existing regulations. This can affect the factor of tax avoidance.

Consensus Theory
Consensus theory is a social theory. Society is a series of events that are close together in the social order. A person is relied on as a personal perception (Martin, 1958). A decision in a company is the result of a joint decision between managers and company leaders. Decisions are the result of communicating with several people. According to Cushman and Whiting (1972), communication is a beneficial and important activity based on an agreement. In this research case, communication between the manager (agent) and leader (principal) requires decision making for the realization of company goals. These different decisions must be made as a joint decision so that the company's performance runs well.

Tax avoidance
Tax avoidance is an effort to reduce the tax burden by not violating tax laws. Companies avoiding tax or fighting against taxes are obstacles that occur in tax collection, resulting in reduced state cash receipts. Tax avoidance carried out by taxpayers does not conflict with applicable laws or legal provisions (Putra & Merkusiwati, 2016). According to Meidawati and Azmi (2019), the higher the level of awareness owned by the taxpayers running the business, the higher their level of compliance.

Capital intensity
According to Budianti and Curry (2018), intensity capital is the wealth of a company in the form of fixed assets. The higher the company has capital intensity, the depreciation expense also increases. This increased depreciation expense can reduce the tax burden owed. Capital intensity carried out by the company must be planned so that the company's profit does not decrease.

Inventory intensity
Inventory Intensity is a measurement of the level of inventory as an investment made by the company. According to PSAK No. 14 regulates the costs arising from the ownership of large inventories must be excluded from additional costs for the existence of large inventories causing reduced profits to the company. If the profit charged decreases, it will result in reduced taxes paid by the company. Usually, these costs cannot be used as a tax shield because they are not a deduction for taxable income (Dwiyanti & Jati, 2019).
Manufacturing companies consist of 3 sectors such as the basic industrial sector and chemicals, various industrial sectors, and the consumer goods industry sector. Each sector has different company characteristics and types of inventory. Measurement of inventory intensity in these 3 manufacturing sectors can be divided into 3 parts, namely slow, medium, and high inventory intensity. This division is done because the manufacturing company has 3 sectors. Inventory intensity can be seen from the average inventory turnover of each sector. Sectors that have a turnover of goods in small companies are categorized in slow inventory intensity and vice versa.

Auditor industry specialization
An audit is an examination conducted to evaluate an organization's system processes and products. This audit is conducted by a competent person and must not be impartial. The person who audits is called an auditor. The purpose of the audit is to find out that the company's financial statements have been prepared based on applicable regulations and in accordance with PSAK standards. The external audit that has tax professional services will support clients to carry out profitable tax planning and not get out of the bounds of taxation rules. This can make companies (clients) avoid high taxes (Setiyawati & Rohman, 2015). According to Sanjaya (2017), there are two approaches to identify specialized industry auditors such as a market share in the industry and a market share in KAP.

Data Collections
The population used in this study are manufacturing companies listed on the Indonesia Stock Exchange for the period 2015-2018. The reason for choosing this population is that the manufacturing sector is a source of economic growth in Indonesia and contributes greatly to tax revenue. This manufacturing sector also has fixed assets and large inventories, where tax avoidance can occur to manage company profits so that they do not decrease. In this study, the sample was determined using the pool sampling method. Pool sampling is a sampling technique that takes all members of the population as a sample. The criteria for determining the sample are as follows: 1. Manufacturing companies listed on the Indonesia Stock Exchange from 2015 to 2018. 2. Manufacturing companies that were not out or delisted from the IDX during the researcher observation period. 3. Manufacturing companies listed on the Indonesia Stock Exchange that have presented audited financial reports as of December 31, 2015, to December 31, 2018. The company has complete data information needed regarding tax avoidance, capital intensity, inventory intensity, and audit industry specialization.

Dependent variable
Tax avoidance or resistance by companies to taxes are the obstacles that occur in tax collection, resulting in reduced state cash receipts. In this study, tax avoidance uses the proxy Books Tax Difference (BTD). The company will try to report high accounting profit for the benefit of shareholders, but carry out strategies so that taxable profits are low. The BTD formula was adopted from research Comprix et al. (2011)

Independent variables
Capital Intensity Capital intensity can be seen from how much the company invests in its fixed assets. Fixed assets owned by the company can reduce the tax value from the results of depreciation or depreciation expenses (Anindyka et al., 2018). In this research hypothesis, Capital Intensity has a positive effect because the asset depreciation method is supported by tax rules. Calculating capital intensity by comparing fixed assets such as equipment, machinery, and various other properties to total assets (Puspita & Febrianti, 2017). H1: Capital intensity has a positive effect on tax avoidance.

Inventory Intensity
Inventory intensity is a measure of the amount of inventory that can be invested by a company. Additional costs arising from investing in inventories must be excluded from the cost of the inventory and recognized as an expense in the period in which the costs are incurred. By removing additional costs from inventory and recognized them as an expense in the period in which the costs are incurred, it can cause a decrease in the company's profit. According to Dwiyanti and Jati (2019), the higher the intensity of the company's inventory, the more tax avoidance actions the company takes will increase. This inventory measurement is used in 3 subsectors in manufacturing companies. This variable uses a control variable, namely the industrial sector. H2: Inventory intensity has a positive effect on tax avoidance.

Auditor Industry Specialization
Auditor industry specialization is the expertise possessed by auditors to audit the financial statements of audited companies. The auditor's expertise used in this study is tax expertise. Good audit quality is related to the specialty of the industry auditor because an auditor will ask the company's management to do the tax by applicable regulations. In this research hypothesis, specialized industry auditors can affect tax avoidance because an auditor who specializes in auditors can assist companies in minimizing the tax expenses. This auditor specialization variable uses a dummy variable, namely the value of 1 if the market share of tax expertise from KAP is equal to or more than 30% and is worth 0 if the opposite (Setiyawati & Rohman, 2015). H3: Specialized industry auditors can affect tax avoidance.

Variable control
Sub-sector is the determination of the industrial sector in tax avoidance, which can be determined because the inventory turnover is slow so that unsold goods require additional costs including maintenance or warehouse costs (Astuti & Aryani, 2017). This study, using three sectors by looking at the average inventory intensity of each manufacturing sector. These three sectors are grouped into slow, medium, and high subsectors. The slow sector is a manufacturing company that has a slow inventory intensity and the types of inventory are mostly in-process and finished goods. There are indications that this subsector will incur additional costs and will reduce the tax burden. The second subsector is the sector with moderate supply intensity. The type of supply in this sector is the inventory of raw materials or raw materials that will be sold for a profit. The last sub-sector is the sector with high supply intensity. This manufacturing company which is classified as a high sub-sector has an inventory of finished goods that are ready to be sold. Manufacturing companies with high subsectors incur small additional costs so that companies do not use inventory intensity for tax avoidance. This variable uses the measurement of sector 3 and sector 5 dummy, where sector 3 is medium inventory turnover, sector 4 is low inventory turnover, and sector 5 is high inventory turnover. In this study, there are more than two categories, so the dummy variable that is formed must be n-1. The measurement used is the sector 3 dummy with medium sub-sector and sector 5 dummy with high sub-sector. Sector 3 dummy by giving a value of 1 if the company is included in sector 3 (medium) and those belonging to other sectors will be given a value of 0. Measurement of inventory Sector 5 dummy, if the company enters sector 5 (high) will be given a value of 1 but if the company is not part of sector 5 will be assigned a value of 0.

Analysis Method
The analysis in this study uses descriptive statistical analysis, panel data, classical assumption test, and hypothesis testing with multiple linear regression analysis as the analysis technique. This study uses Microsoft Excel 2010 and STATA MP version 13 as data processing programs. The model used in this research is as follows: In this study, a determination test was used to test the ability of the research model to show variations in the dependent variable. The value of the coefficient of determination is between 0 and 1. If the value 0 of R 2 shows a small value, it means that the ability of the independent variable to explain the dependent variable is limited, whereas if R2 is close to the value of 1 it means that the independent variable provides almost all the information needed by the dependent variable (Ghozali & Latan, 2015).

Partial test (T-test)
Hypothesis testing can be partially tested using the T-test formula. The purpose of the T-test is to test whether or not the effect of each independent variable (X) on the dependent variable (Y) is present. The t-test shows how far the influence of one explanatory or independent variable individually in explaining the variation of the dependent variable.

Results and Discussion
The research object used in this study is a manufacturing company consisting of the basic industry and chemical sector, various industries, and consumer goods industries listed on the Indonesia Stock Exchange (IDX) in the 2015-2018 period. Based on information from Table 1, the number of companies being sampled is 88 companies with a 4-year observation year, so that the number of companies is 352 manufacturing companies studied. However, there are 20 companies affected by outliers because the data is considered abnormal to be studied, so the final sample is 83 companies with a total sample of 332 manufacturing companies used in this study. Table 2 shows the results of the descriptive statistics in this study with a sample size of 332 for each variable. This sample is obtained from annual reports, annual financial reports of manufacturing companies for 2015-2018.
The dependent variable in this study is tax avoidance which is proxied by BTD. In Table 2, the minimum value of the BTD variable is -0.1739036 or -17.39%, and the maximum value of the BTD variable is 0.1548338 or 15.48% obtained. The mean value of the BTD is -0.0018477 or -0.18%, so it can be interpreted that the manufacturing company in this study has a low accounting profit compared to fiscal profit and the average manufacturing company in this study does not do tax avoidance. The BTD variable has a standard deviation value of 0.0366439 or 3.66% and it can be interpreted that tax avoidance deviates by 0.0366439 from a mean value of -0.0018477.

Analysis of Classical Assumption Test Data
The normality test in this study used the Skewness and Kurtosis test between -2 to +2 to determine the normality test. Based on the results of the normality test, all variables have met the requirements for normal distribution so that the data has been normally distributed.
The multicollinearity test aims to determine whether the regression model found a correlation between the independent variables. A good regression model should not correlate with the independent variables. All variables in this study have met the requirements to be free from multicollinearity with a tolerance value of more than 0.1 and a VIF value of less than 10. So that this study does not have multicollinearity problems.
Heteroscedasticity test in this study using the random effect model estimation. Where the regression estimation of the random effect model does not experience heteroscedasticity because the coefficient used is general least square. These results indicate that the panel model is homoscedastic, which means the variables are not affected by heteroscedasticity.

Data Model Feasibility Test
Based on Table 3, the chow test, hausman test, and breusch and pagan langrangian multiplier test. After the panel data estimation tests were carried out, this study used a random effect estimation model. R squared shows a value of 0.0546 or 5.46%. This shows that the independent variable is able to explain the dependent variable by 5.46% while the remaining 94.54% is explained by other variables outside of this study. Testing the hypothesis of this study using multiple linear regression test using panel data, the summary of the test is in Table 4.  Table 4, for the number of data (n) 332 with the number of variables (k) of 5, it can be obtained that the number of T table is (a/2; nk-1 or df residual) where (0.1/2; 332-5-1) then T table of (0.05; 326) is 1.649541.
Based on the results of statistical tests for the intensity capital variable, the T value is -0.95 with a significance level of 0.340, because T test<T table (-0.95<1.649541) and the significance level is 0.340> 0.1. So it can be concluded that H1 is not supported, capital intensity has no significant effect on tax avoidance. Furthermore, H2 is rejected, it means the inventory intensity has no significant effect on tax avoidance. On the other hand, H3 is supported which means that the specialization of industrial auditors has an effect on tax avoidance.
On the other hand, in Dummy Sector 3 the significance level is 0.039 <0.10. So it can be concluded that DS3 has a negative effect on tax avoidance. Furthermore, DS5 has a negative effect on tax avoidance.

Discussion
The effect of capital intensity on tax avoidance Based on the tests that have been conducted, this study rejects Hypothesis 1 which states that capital intensity has a positive effect on tax avoidance. This study proves that capital intensity does not have a significant effect on tax avoidance because large manufacturing companies usually have used fixed assets that have expired. The results of this study are supported by research Budianti and Curry (2018), and Puspita and Febrianti (2017), because companies that have large fixed asset ownership of manufacturing companies, the indications for tax avoidance are getting smaller and there are differences in the useful life of a company. fixed assets of the company and taxation and allowed to depreciate their fixed assets. Based on the consensus theory, manufacturing companies as corporate taxpayers have been responsible for their obligation to pay taxes to the state and an increase in tax policy policies on tax avoidance. Based on the average value of CAPIN of 38.64%, manufacturing companies in Indonesia are still low in investing in fixed assets so that managers do not choose the capital intensity method in avoiding taxes.

The effect of inventory intensity on tax avoidance
Based on the tests that have been conducted, this study rejects Hypothesis 2 which states that inventory intensity has a positive effect on tax avoidance. This study proves that inventory intensity has no significant effect on tax avoidance because manufacturing company investment in inventory has no impact and is not the right way to reduce tax. This research is supported by Lestari (2015) because it is not the right way to do tax planning with the inventory investment method, and the amount of inventory owned by the company is not a factor in determining the size of the tax that must be paid by manufacturing companies. . The levels and types of supply in sector 3, sector 4, and sector 5 vary each year and have different policies in investing in their supplies. Manufacturing companies have carried out their responsibilities as taxpayers because they follow the tax policies implemented in Indonesia.

The Influence of Auditor Industry Specialization on Tax Avoidance
Based on the tests that have been carried out, this study accepts Hypothesis 3 which states that industrial auditor specialization has an effect on tax avoidance. This study proves that the specialization of industrial auditors has a positive effect on tax avoidance because the data shows that 13% of auditors in sample companies have specialties, and auditor specialization is an indicator that the quality of auditors in auditing financial statements is good so that they are able to provide an assessment that the company's accounting reports have been prepared fairly. This is what encourages managers to increase the intensity of tax avoidance by optimizing loopholes in the rule of law to obtain maximum tax savings. Industry-specialized auditors do not guarantee that a manufacturing company does not evade tax. According to (Lee & Kao, 2018), a public accounting firm can work with a company because when auditors have a better understanding of the industry characteristics of audit clients, they will improve the quality of audit clients' earnings than nonindustrial specialist audit firms. Auditors who collaborate with companies can risk their independence as auditors and if caught by the tax officer it will make the KAP (Public Accountant Office Branch) look bad.

Conclusion
Based on the objectives and research results, this research was conducted to obtain statistical evidence of tax avoidance in the Indonesian manufacturing industry on the Indonesia Stock Exchange in the 2015-2018 period. Capital intensity does not have a significant effect on tax avoidance because large manufacturing companies usually have used fixed assets that have expired. Manufacturing companies also cannot do tax avoidance because of the tax policies that are continuously updated by the government. Inventory intensity does not have a significant effect on tax avoidance because it is because manufacturing company investment in inventory does not have an impact on tax reduction. Furthermore, the auditor industry specialization has an effect on tax avoidance because 13% of auditors of the sample companies have specialties, and auditor specialization is an indicator that the quality of auditors in auditing financial statements is good so that they are able to provide an assessment that the company's accounting reports have been prepared fairly. This is what encourages managers to increase the intensity of tax avoidance by optimizing loopholes in the rule of law to obtain maximum tax savings.
This study has limitations when processing data such as many companies that do not provide the complete data needed by researchers in this study, and this study is limited to only considering the control variables of the industrial sector, many other factors that can be an influence for corporate tax avoidance. From the existing limitations, suggestions for further research are to use other variables that can affect tax avoidance such as profitability, firm size, leverage, and so on.