Role of Audit Committee in Tax Avoidance of Family and Non- Family Firms: Evidence from Indonesia

Companies often use tax avoidance and cost of debt substitutionally to achieve the desired marginal tax rate, without attracting the attention of the Tax Department of the State. We examine the role of the audit committee in supervising and reducing the trade-off practices of 222 public companies with 1.101 observations in Indonesia from 2011 – 2015 (before the tax amnesty program in 2016). We also examine the structure of company ownership (family and non-family firms) determines the effectiveness of the role of the audit committee in reducing the trade-off between tax avoidance and cost of debt. Data were analyzed using multiple regression data panel with pooled least square. Proxy of tax avoidance in this study uses Lim (2011), while the indicator of the audit committee uses the proportion of the audit committee compared to the number boards of commissioners, in dummy variable. The results suggest that the role of the audit committee is moderated the trade-off between tax avoidance and cost of debt, but in a contrary way. The role of the audit committee is not proved to reduce the trade-off practices, on the contrary increase the cost of debt which can result in reducing tax payments and increasing tax risks. After the advanced analysis, we found evidence that the role of the audit committee is stronger in non-family firms than in family firms. The moderation effect of audit committee found not significant in family companies, though have a significant effect on non-family companies.


INTRODUCTION
Companies often use tax avoidance and cost of debt substitutionally to achieve the desired marginal tax rate, without attracting the attention of the Tax Department of the State, by minimizing the possibility of higher tax risk (DeAngelo and Masulis, 1980).This raises an opportunity for management to do moral hazard, consequently, the reliability and transparency of the company's financial statements will be sacrificed.Another negative impact is that companies that do the trade-off practice are viewed by the public as riskier.Hanlon & Slemrod (2009) find news related to a company's tax avoidance causing the stock market to react negatively to the news.DeAngelo and Masulis (1980) also found that as the income tax rate increased, the company will substitute with a more significant cost of debt.But companies with lower tax rates tend to lower the amount of debt, so the cost of debt is also reduced.Graham & Tucker (2006) found companies that involved in tax avoidance experience decreasing in debt or have the cost of debt three times smaller than other companies.This finding supported Lim (2011) also found that most companies in Korea substitute tax avoidance and cost of debt.Lim (2012) who examined public companies in Korea in 2000 -2006 found that tax avoidance negatively affects the cost of debt, and the substitution effect is stronger in conditions where taxation sanctions are more severe.
Companies in Indonesia appropriate used as the object of this research because of several conditions: (1) the rate of corporate tax is more than double the bank credit rate.If the corporate income tax rate since 2010 is 25% flat rate, while the interest rate of the corporate bank is 10.25% -11.50% (Indonesia Investment, 2016).Significant tariff differences between corporate income tax and cost of debt result in more significant possibility of substitution between tax avoidance and cost of debt; (2) the Indonesia's banking policy to lower bank loan interest rate from two digits to single digits at the beginning of 2016, widening the difference between the income tax rate and cost of debt.This is not immediately followed by the decrease of company's income tax rate, which is planned to be reduced from 25% to 17% (Indrastiti, 2016); (3) Enforcement of tax laws in Indonesia is ongoing and restrict the opportunity of tax evasion in Indonesia.In Indonesia sanctions of the tax increases of up to 150% of unpaid taxes (Act KUP No. 28 of 2007 Article 13 paragraph 3).Under conditions of a country with strict tax rules, the trade-off between tax avoidance and cost of debt will raise the risk of the companies.
The second objective of this study is to examine the role of the audit committee as a moderator of the trade-off between tax avoidance and cost of debt.Research conducted by Desai and Dharmapala (2009) argues that tax avoidance led to opportunistic actions of managers and this factor was increasing in companies with low corporate governance.Izma (2013) wrote the audit committee as the key to effective corporate governance, particularly in public companies.In the analysis of the effect of tax avoidance on the cost of debt, the audit committee is expected to moderate this relationship.The more effective of the performance of the audit committee, the trade-off between tax avoidance and cost of debt will be weakened.According to the Decree of the Chairman of the Indonesia Capital Market Supervisory Board of KEP-643 / BL / 2012, the responsibilities of the audit committee shall, among others, review the compliance with prevailing laws and regulations in the company's activities, reviewing the activities of risk management performed by the Board of Directors.It is seen that the audit committee is responsible for reducing the trade-off practices in the company, therefore the tax risk of the company even decrease.
Nevertheless, the structure of company ownership determines the effectiveness of the audit committee to mitigate the adverse effect of trade-off practices.Izma (2013) argues that audit committees in family-dominated ownership firms are more likely to lose their voting power and independence than companies with publicdominated ownership structures.The audit committee's role is allegedly stronger in non-family firms.The indicator used to measure family ownership is the percentage of ownership controlling families.Controlling families is determined by looking at the most significant shareholder percentage, at least 10% ownership (Faccio and Lang, 2002).Indonesian public companies dominated by family companies.Carney & Hamilton-hart (2015) conducted a study on the structure of ownership of Indonesian public corporations in 1996 and 2008 found that the ownership structure of Indonesian firms is still dominated by families with a very high percentage of ownership.The dominance of the family in the ownership structure of the company results in the "dwarfing" of the audit committee's role in improving corporate compliance, particularly with regard to tax regulations.Audit committees, especially in Malaysia, also face challenges related to transparency of financial statements, litigation risks, pressure from shareholders especially public companies whose majority shareholder is family (Izma, 2013).
The first contribution of this study is the first study using the audit committee as a moderator that moderate trade-off between tax avoidance and cost of debt in Indonesia.The audit committee is expected to be an agent that can reduce trade-off practice in Indonesia in the long term.The second contribution is the first study to reveal that effect moderation of audit committee is higher in firms with a smaller percentage of family ownership (non-family ownership).

Effect of Tax Avoidance on Cost of Debt (Trade-off Practice)
According to Schallheim and Wells (2008), firms prefer tax avoidance over cost of debt because (1) tax avoidance cheaper than cost of debt, (2) credit agreements often need requirements that companies must meet, such as asset security, cash restriction, and dividend restriction, therefore the cost is higher than tax avoidance, (3) tax shield exploit provision in the accounting rules that allow the firm to reduce the tax without affecting the income statement.In the high tax law enforcement condition, the company will reduce tax avoidance and increase the cost of debt, to minimalization tax risk.According to KPMG (2007) tax risk is the risk of noncompliance with tax rules, combined with unintended tax consequences of company transactions.Elgood et al. (2004) divided tax risks into two categories of internal and external tax risk.External tax risk occurs through new tax laws, regulatory and legislative changes that sometimes are hard to manage by companies.Internal tax risk classified as transactional, operational, compliance, and financial accounting risks (Erasmus, 2014).Tax avoidance measurements in this study use proxies in Lim (2011) which uses book-tax difference (BTD) and discretionary accrual (Dechow et al., 1995).The indicator of tax avoidance (Lim, 2011) limits the tax avoidance only on BTD which cannot be explained by the discretionary accrual.The research that has been done by Graham & Tucker (2006) and Lim (2011) shows that tax avoidance reduces the tendency of companies to owe.Tax avoidance has an adverse effect on the cost of debt or support trade-off theory.The greater tax avoidance, it will reduce the cost of debt, so the hypothesis proposed as follows: H1: Tax Avoidance negatively affects the cost of debt

The Audit Committee as Moderator to The Trade-Off between Tax Avoidance and Cost of Debt
Corporate governance arises by reason of the separation between ownership and corporate management that can cause agency problem.One of the pillars of good corporate governance is through an audit committee.The Audit Committee is a committee established by and responsible to the Board of Commissioners in assisting with the duties and functions of the Board of Commissioners (Decision of the Chairman of Bapepam and LK Number: Kep643 / BL / 2012).One of the tasks of the audit committee is to review the financial information to be issued to the issuer or public company to the public and/or the authority, among others, financial reports, projections and other reports relating to the issuer or public company's financial information.The audit committee has a very important and strategic role in the company's oversight mechanism.Audit Committee is one of the company's organs that oversees the effectiveness of corporate governance.Anderson et al. (2004) prove that the size of the audit committee is negatively related to the cost of debt.The existence of the audit committee is expected to give a positive effect to the company which is the trust of the creditors, therefore provide a lower level of debt costs.In large companies, the audit committee is required to comprise between three to six members (Burke & Guy (2002) in Ferreira (2008)).KPMG and the Institute of Director's Audit Committee Forum (KPMG's ACF, 2006) proposed a guideline, that committee should be large enough to represent a balance of views and experience, but small enough to operate efficiently.According to the Indonesia regulation, the minimum number of audit committee members is three people.H2: The audit committee weakens the trade-off between tax avoidance and cost of debt Caselli and Gennaioli (2003) and Burkart et al. (2003) state that ownership concentration empowers family members to achieve their goals better than other shareholders.Family control may eliminate agency problems from the conflict between shareholders and managers.Family members tend to have a longer horizon investment compared to other investors (James, 1999;Stein, 1989).According to Demsetz (1983), family firms may choose non-monetary benefits and remove resources from profitable projects consequently damage firm performance.Because families are motivated to pass on their ownership to future generations (Casson, 1999;Chami, 2001;James, 1999), they may act to reduce the risky capital structures or investment projects.

Moderation Effect of Audit Committee on Trade-off Practice in Family Vs Non-family Firms
As we know about family firms and the effect on the effectiveness of audit committee, we suspect that the role of the audit committee in family firms is going to be less dominant than in non-family firms.Rather than experiencing the traditional manager-owner conflict, the conflict between shareholder groups is the prevalent agency problem in East Asian companies as compared to the US and UK (Shleifer and Vishny, 1997;Claessens et al., 2000Claessens et al., , 2002)).Members of the controlling family can exercise control over the board (Anderson et al., 2004) which in turn may provide them with opportunities to expropriate minority shareholders.H3: Audit committee moderation on trade-off is stronger in non-family companies than in family companies

RESEARCH METHODOLOGY
The model in this research is modification of Lim (2011) model by adding audit committee variable (AC) as moderator and family variable (Fam) as the independent variable.The cost of debt indicator in this study uses the ratio of cost of debt to the average short-term and long-term debt that resulted in a cost of debt (Lim, 2011).The audit committee's effectiveness proxy uses a ratio of the number of audit committees to the number of commissioners.
The family indicator in this research use dummy variable that is number 1 if controlled by family and number 0 if not controlled by the family.The criteria of a company controlled by the family are (1) one of the shareholders and the board of directors is held by the same family, judging by the name of the family, (2) controlling shareholders are individuals or private, non-listed companies (Faccio andLang, 2002 andMaury, 2006) with minimum ownership of 10%, (3) to identify the tax heaven company which is actually a non-listed firm, is viewed from the shareholder's name or whether the directors are held by foreigners.When held foreigners, it includes non-family firm, if a not occupied foreigner, then the family firm.
The control variable in this study is the company's age listing on BEI, leverage, cash flow from operation, company size, plant and equipment property, and negative equity.The following are the model equations formed in this study: In this study, a proxy of the audit committee is a dummy variable.I measured the effectiveness of audit committee using a proportion of audit committee compare with numbers of the board of commissioners.If the proportion is more than 1, then the value is 1, otherwise, the value is 0.

Sample Determination
The population of this study are companies listed on the Indonesia Stock Exchange in 2011 -2015.Sample selection criteria used are: 1.The company does not include financial institutions.
2. The company excludes construction and property services firms which income tax is not based on net income, therefore it is not relevance with the tax avoidance indicator in this study.3. The company has complete data for at least 4 years.

ANALYSIS AND INTERPRETATION
Based on the predetermined criteria, the number of samples used are 222 companies from totals of 383 companies.The total sample in this study is 1110 observations obtained from 222 companies for five periods (2011-2015).However, due to the process of data reduction, outlier, therefore, removed 9 observations, and produce total data 1101 observations to be analyzed.Table 5 shows the result of regression testing for both models.The result of regression test of model 2 as a whole shows tax avoidance has a negative effect on the cost of debt in Indonesia (significant at 5%), meaning that the company practices trade-off between tax avoidance and cost of debt in tax avoidance framework, therefore H1 is received.These results support the Lim (2011) study that proves tax avoidance lowers the cost of debt of firms through trade-off practices.The results of moderation testing found that tax avoidance and audit committee have a positive effect on the cost of debt, meaning that tax avoidance and audit committee increase cost of debt.The results are shown in Table 3 where TA*AC moderation is significant at the 1% level.This result did not prove the hypothesis (H2), because of the better quality of corporate governance, it proved unable to reduce (weaken) trade-off between tax avoidance and cost of debt (H2 is rejected).The audit committee's moderation effect causes tax avoidance and cost of debt to be positive, meaning that when tax avoidance is high, and the audit committee's proportion is high, the cost of debt becomes higher.The higher cost of debt gives an indication of the greater deductible expense deducted in the tax financial statements, thus reducing the tax payable.The greater proportion of audit committees to the board of commissioners leads the audit committee to have greater "power" to influence decisions in shareholder meetings, including decisions relating to tax avoidance.
The positive effect of the moderation tax avoidance-audit committee on the cost of debt is tested again with graphs to determine the positive influence of both variables.The results of the moderation effect are shown in Figure 1.The blue line in Figure 1 shows the increase in tax avoidance (TA) resulting in a lower cost of debt (COD).The green line shows an increase in TA and audit committee, resulting in higher COD.There are several reasons why this hypothesis is rejected.First, the audit committee in Indonesia views tax avoidance and cost of debt tools as an efficient perspective for the company as a taxpayer.The utilitarian approach to tax avoidance states that to determine whether the right or wrong of action needs to measure the number of utilities or disutility to taxpayers, government, and society (Filho, 2014).For taxpayers when using both tools, the utility value increases, moreover the company's performance looks better than last year.On the other hand, it is also necessary to evaluate whether these practice increases or decreases the utility of the state and the society.These schemes can reduce the value of government and community utilities if causes state revenues to fall and the government cannot provide benefits to the community.However, if the quality of the government is not as expected and the state income is not effectively utilized to increase social utility, then the efficient perspective of these taxpayers can be justified (Filho, 2014).In this case, the quality of public administration and state politics is the main key for the measurement of government and community utilities (Filho, 2014).
Second, the audit committee considers tax risk for companies in Indonesia to be low, thus they do not make this matter as a priority to be assessed or evaluated.This point issues the opportunistic behavior for management.In Indonesia, this view is supported by the survey that was conducted by KPMG (2014).Based on the result of the survey of tax executives in Indonesia, tax risk is not the biggest challenge faced by Indonesian companies.They mention the biggest risk challenges are (1) uncertainty and volatility in economics, regulatory, and political, (2) government regulation or public policy.The survey results show that only 3% of 30 respondents rated tax risk as the company's most significant risk, besides 27% said they needed less time to discuss tax risk in meetings, while 33% said they did not need additional time to discuss tax risks (KPMG, 2014).
Third, some people view tax avoidance as a natural law or universal law.Deontological approach to tax avoidance states right or wrong an action is measured by norms that are trusted by society, meaning that tax avoidance can be a universal law for individual taxpayers (Filho, 2014).However, Prebble and Prebble (2010) stated that universal tax avoidance could not be justified if all parties (taxpayers, government, and society) will experience adverse effects if all people do the tax avoidance.The declining state revenue has caused the state not able to provide benefits to the society who also impact on the creation of new taxes which ultimately have a negative impact on the company as a taxpayer.
In model 2 (Table 5), in addition to tax avoidance and TA*AC moderation, size, cash flow from operations, negative equity, and infrastructure and agriculture industry affect the cost of debt.The TA*AC, size, and negative equity variables are the three variables with the most substantial influence on the cost of debt (significant at 1%), and the TA*AC moderation variable has the most substantial influence on the cost of debt.This shows that the moderation of audit committee impacts strongly on tax avoidance and cost of debt relationships.
Figure 2 (Appendix 1) shows the moderate effects of the audit committee on each industry sector.This study divided industry sectors into seven sectors, consumer goods, basic and chemical, infrastructure, retail and services, mining, agriculture, and others sectors.Based on the moderating effects of all industry sectors in figure 2, it can be concluded that the audit committee moderation effects are strongest in the basic chemical and mining industry sectors, while the weakest effects are agriculture.This means that the audit committees in basic chemical and mining industry sectors have the highest power in influencing management decisions related to tax avoidance compared to other sectors.
The results of regression testing related to ownership structures on the audit committee moderation effects can be seen in table 6.It shows that trade-off practices are not found in companies with family ownership, otherwise found in non-family enterprises with a significance of 5%.The moderation effect of audit committees on tradeoff practices is also not found in family firms, but significant in non-family enterprises (significant at 1%).Nonfamily companies in this study dominated foreign companies, with the composition of 75 foreign companies and 3 government companies.This finding is consistent with Casson's 1999 study; Chami, 2001; James, 1999, who found family companies more intergenerational stewardship-oriented and more risk-averse than non-family firms.James (1999) and Stein (1989) found family companies tend to make long-term investments compared to other types of investors.Demsetz (1983) found that family firms prefer to avoid profitable projects that can destroy company performance.A strong vision of a family firm leads to the insignificant role of the audit committee, especially in the practice of trade-off (Table 6).The biggest decision is for the family members who are leaders of the company.

2 )
Measure Tax Avoidance, Cost of Debt, Audit CommitteeTax avoidance measurement uses the tax avoidance measurement scale used byLim (2011).The first step of tax avoidance measurement is to find the value of discretionary accrual with modified jones(Dechow et al., 1995).The residual value of model 1 is Discretionary Accrual (DA).Accrual it : total accrual, whereas Accruals= Net Income -Cash Flow from Operations ΔREV itThe second step is to seek tax avoidance(Lim, 2011) which is a representation of tax avoidance.How to separate the book-tax-different components that are not caused by earnings management, and identify these components as tax avoidance.The residual value of model 2 is tax avoidance indicator.

Figure 1 .
Figure 1.Moderation Effect of Audit Committee on Tax Avoidance (TA) Effect on Cost of Debt (COD)

Table 3 .
Variable Descriptive Statistics

Table 3
presents the descriptive statistics of variables.It shows that on average COD is 0.091, TA is 0.001, moderation TA*AC is 0.024.The standard deviation for all variables shows that COD is 0.062, TA is 1.056, TA*AC is 0.201.

Table 4 .
The result of Pooled Least Square (PLS) Model 1 and Model 2 Based on the test results of both models in table 4, shows the result that model 1 has R Square of 9.5%, with a significance level of 0.000, meaning that model 1 meets the goodness of fit test.The result of testing model 2 has R Square equal to 10,7% with significance level equal to 0.00, and then model 2 also fulfill the goodness of fit.The comparison of R Square of both models shows that model 2 has a higher R Square, meaning the ability to explain predictor to criterion variable in model 2, higher than model 1.It shows the audit committee as the moderating variable in model 2 can improve the prediction result of the cost of debt.
Pooled Least Square (PLS), Fixed Effect (FE) or Random Effect (RE).Chow-test testing is performed to determine whether the regression model is Pooled Least Square (PLS) or Fixed Effect (FE).The test results indicate that this research model is PLS.The Lagrange Multiplier (LM) test also strengthens the Chow-test test results that the model includes PLS.The determination of this regression model is done by Stata program.

Table 6 .
Anderson et al., 2004; Shleifer and Vishny, 1997; Claessens et al., 2000 also support the results of this study.Output Regression for Family VS Non-Family OwnershipFigure 2.1 Consumer Goods Industry