1. Introduction

Nowadays, the concept of modern risk-oriented auditing has been introduced with the international auditing standards and the new auditing standards promulgated in 2006. In the case of internal control failure such as Enron and WorldCom, the modern risk-oriented auditing has gradually become the trend of the entire auditing industry, which requires auditors to make a reasonable risk assessment on the enterprise to be audited. Tax avoidance reduces the transparency of enterprises (Chen Dong and Tang Jianxin, 2012) [1] and weakens the quality of accounting information (Weber, 2009) [2] . As a result, auditors will face more audit risk and have to make more efforts. Based on the “deep pocket” theory, certified public accountants would correspondingly raise the pricing of audit services. We wonder that whether tax avoidance will arise higher audit fees. In addition, the property rights of different enterprises may have an impact on the relationship between tax avoidance and audit fees. On one hand, compared with non-state-owned enterprises (non-SOEs), state-owned enterprises (SOEs) have a social goal of paying more taxes to help the government except the goal of “maximizing profits”. Thus, the level of tax avoidance in state-owned enterprises is relatively low (Wang et al., 2010) [3] . On the other hand, for the natural connection between state-owned enterprises and the government, the government provided the invisible guarantee for the state-owned enterprises which will also reduce the litigation risk related to the financial report.

Since Simunic (1980) [4] first started to study the determinants of audit pricing, audit pricing has become a topic of general concern to scholars. Factors affecting the audit pricing in prior research can be divided into the micro factors and the macro factors. Micro factors include the characteristics of the firm, the characteristics of accounting firm, the characteristics of the management, the auditor’s characteristics, the risk factors and the behavior of corporates. Macro factors mainly include the law and the economic environment. Specific is shown in Figure 1.

However, prior research on the factors affecting audit fees has paid little attention to the impact of the behavior of enterprises, such as tax avoidance. To fill this void in the literature, this study mainly focuses on the influence of tax avoidance on audit pricing, and combine with the basic characteristics of the enterprise-property rights to make further research.

This paper selects the A-share non-financial listed companies in Shanghai and Shenzhen Stock Exchanges from 2012 to 2015 as a sample, explores the relationship between tax avoidance and audit fees, and introduces property rights of enterprises to further explore its impact on tax avoidance and audit fees. The

study found that the audit fees and the degree of corporate tax avoidance is positively correlated. At the same time, compared to state-owned enterprises (SOEs), the greater the degree of avoidance in non-state-owned enterprises (non-SOEs), the higher the audit fees charged by accounting firms.

Our paper makes several contributions. On one hand, this paper studies the impact of tax avoidance on audit fees, not only will this article enrich the research on the factors affecting the pricing of auditing, but also show property rights have a significantly different impact on tax avoidance and audit fees. On the other hand, this paper may be of some reference to regulators and audit pricing.

2. Theoretical Analysis and Research Assumptions

2.1. Tax Avoidance and Audit Fees

Auditing pricing is a function of the auditor’s efforts and potential losses such as litigation risk, loss of reputation, regulatory penalties (Simunic, 1980) [4] . Tax avoidance enhances the auditor’s efforts and potential losses in at least three ways:

Tax avoidance reduces the quality of accounting information. When it comes to tax avoidance how to affect the quality of accounting information, the academic view is mainly concentrated in earnings management. Many tax-related accruals contribute to earnings management (Hanlon et al., 2012) [5] . The impact of earnings management on the information about future cash flows in tax-related accruals increases with the likelihood of financial misstatement and the company facing litigation (Marry et al., 2007) [6] . Daliwal et al. (2010) believe that since income tax is the last item to be identified in the financial report, it will also be the last opportunity for earnings management [7] . Managers will influence the effective tax rate of the enterprise through accounting items such as deferred income tax impairment or accounting―tax differences to achieve the purpose of earnings management. Deferred income tax expenses play an important role in the total accruals and unconventional accruals in earnings management to avoid falling earnings. When tax avoidance becomes a tool for companies to conduct earnings management, they will inevitably have an impact on the quality of earnings. In summary, corporate tax avoidance reduces the quality of accounting information through earnings management, so that the audit risk faced by auditors increases.

Tax avoidance reduces the transparency of corporate information. Tax avoidance itself is not transparent, it is a game between enterprises and tax authorities, companies will hide their tax avoidance behavior as much as possible to avoid the tax authorities supervision, accounting and accounting adjustments will reduce the transparency of information. At the same time, the tax planning strategy alters the company’s organizational structure and increases the complexity of financial activities and organizational structures. The complexity of organizational structures and tax avoidance transactions may not be passed on to outside investors and analysts (Weber, 2009) [4] . Corporate tax avoidance has also changed the distribution and direction of capital and assets. If the distribution and flow of capital and assets are difficult to be used by external investors to understand the origin and sustainability of surpluses and cash flows, the transparency of corporate financial reporting and operations will be decreased.

Enterprises with a high degree of tax avoidance may have more serious agency problems. Computation of income tax costs is complex, the recognition of income tax-related accruals need to use professional judgments, increasing the information asymmetry between senior management, shareholders and auditors, which provides management with damage to the interests of shareholders and expand their own earnings opportunity. Previous research shows that some managers can use complicated tactics to conceal their conspiracies of rent extraction for tax-saving (Desai and Dharmapala, 2006) [8] . The extent of rent-seeking is affected by the complexity of the tax-avoidance transactions. The more complicated and concealed the design of tax avoidance transactions, the greater the management’s chances and rents for opportunistic private gain. In turn, rent-seeking may further increase the willingness and behavior of management to implement tax avoidance transactions, resulting in a further increase in tax avoidance. Management diverted corporate resources using tax avoidance transactions that ultimately distorted the company’s financial reporting. Share prices decline as companies disclose tax avoidance [5] . Corporate tax avoidance is even linked to stock price crash (Kim et al., 2011) [9] . The loss of shareholders will bring litigation against companies and auditors.

Based on analysis above, the tax avoidance of the company affects the audit risk from at least 3 aspects such as the quality of accounting information, the transparency of enterprise information and the management rent-seeking. It also raises the potential risks and potential losses of the audit failure, including litigation risk, reputation Loss, regulatory penalties and so on. According to the “deep pocket” theory, the increased audit risk will lead to higher audit fees. Therefore, this paper proposes hypothesis one:

H1: The degree of corporate tax avoidance is positively related to audit fees.

2.2. Tax Avoidance, Property Rights and Audit Fees

The differences between state-owned enterprises and non-state-owned enterprises in respect of motivation for tax avoidance, degree of tax avoidance and bankruptcy risk may lead to difference between the relationship of tax avoidance and audit fees.

From the perspective of tax avoidance motivation and degree of tax avoidance, non-state-owned enterprises have a greater degree of tax avoidance than state-owned enterprises. On one hand, in fact, taxation is a game between the enterprise and the taxation department. However, for the state-owned enterprises, the government is the beneficiary of state-owned enterprises in both profits and tax forms. Moreover, the government needs the state-owned enterprises to pay taxes rather than motivating the controlling shareholders of state-owned enterprises to maximize the wealth of state-owned enterprises (SOEs). On the other hand, the performance of SOEs is related to the promotion of their managerial positions and career. The profits of financial reports are a vital way to demonstrate their performance (Wang et al., 2010) [3] . Wu Liansheng (2009) and Chen et al. (2015) also verified that the degree of tax avoidance by state-owned enterprises (SOEs) is smaller than that of non-state-owned enterprises (non-SOEs) [10] [11] .

Judging from the bankruptcy risk of the enterprises, the bankruptcy risk of the state-owned enterprises (SOEs) is relatively low compared with the non-state-owned enterprises (non-SOEs), which reduce the auditor’s litigation risk. Government support and political connections behind the state-owned enterprises provide them with invisible guarantees. Besides, state-owned enterprises are more likely to receive government aid when their operations are in trouble. To sum up, the invisible guarantee provided by the government to state-owned enterprises reduces the possibility of the auditors experiencing litigation in the future and reduces the audit risk, so the audit fees will drop. Based on this, hypothesis two is proposed below:

H2: The relationship between tax avoidance and audit fees in non-SOEs is more pronounced than in SOEs.

3. Research Design

3.1. Sample Selection and Data Sources

This paper takes A-shares listed companies in Shanghai and Shenzhen stock exchange as a sample. In order to exclude the impact of income tax reform in 2008 on the calculation of MaETR and MaRatediff which are both the 5-year moving average (the 5-year moving average for year t refers to the average from year t-4 to year t.), the sample is selected from 2012 to 2015. All the financial data come from CSMAR database. We screened the samples according to the following criteria: 1) excluding companies in the financial and insurance industry; 2) excluding ST and * ST companies; 3) excluding samples whose total accounting profit is less than 0 and current income tax expense is less than 0; 4) excluding the effective tax rate (ETR) is greater than 1 or less than 0; 5) eliminating missing values of variables. Finally, firm-year observations were remained. In addition, we winsorized continuous variables in 1% and 99%. We collect data and make statistical analysis by Excel 2010 and Stata 11.

3.2. Variables Definition

3.2.1. Measurement of Tax Avoidance

For the measurement of tax avoidance, we adopt several ways: effective tax rate (ETR), the difference between the nominal tax rate and the actual tax rate (Ratediff), moving average of ETR (MaETR), moving average of Ratediff (MaRatediff).

To begin with, effective tax rate denotes the proxy measure of tax avoidance most frequently used by academic researchers. ETR means the amount of taxes per pretax income. The lower ETR, the higher degree of tax avoidance. We follow the research of Wu Liansheng (2007) and Wu Wenfeng et al. (2009) to calculate effective tax rate(ETR) [10] [12] . The formula is as follows:

ETR = (current income tax expense − deferred income tax expense)/(pre-tax accounting income + provision for impairment other than bad debt provisions − investment income + investment income received in cash) (1)

Secondly, we also use the difference between nominal tax rate and the effective tax rate (Ratediff) to measure. The higher Ratediff, the higher degree of tax avoidance.

At last, corporate tax avoidance is an intertemporal decision (Dyreng, 2010) [13] , ETR or tax rate difference can’t measure degree of tax avoidance properly, so this article also uses the 5-year moving average of the ETR (MaETR) and the 5-year moving average of Ratediff (MaRatediff). The 5-year moving average for year t refers to the average from year t-4 to year t.

3.2.2. Control Variables

Reference to initial literature, we employ the following control variables: company size(SIZE), asset turnover(TAT), company’s inherent risk (RISK), return on total assets (ROA), financial leverage (DFL), whether the auditor is a big4 firm or not (BIG4), whether the company reports net loss or not (LOSS), audit opinion (OP). We also introduce the dummy variable of industry (IND) and year (YEAR) in all regressions to control the year and industry’s impact. Definition of each variable is shown in Table 1.

3.3. Empirical Model

Based on the above analysis and related literature, we design the following fixed effects model to test H1 and H2:

$\begin{array}{l}\text{LNFEE}={\beta }_0+{\beta }_1\text{TA}\left(\text{ETR}/\text{Ratediff}/\text{MaETR}/\text{MaRatediff}\right)\\ +{\beta }_2\text{Controls}+\text{IND}+\text{YEAR}+\epsilon \end{array}$ (2)

Consistent with H1, we except the coefficient of ETR and MaETR to be negative significant and the coefficient of Ratediff, MaRatediff to be positive significant.

In order to test H2, we divide the full samples according to different property rights into state-owned enterprises (SOEs) samples and non-state-owned enterprises(non-SOEs) samples. If the coefficients of TA in SOEs are more significant than in non-SOEs, the result is consistent with our prediction.

4. Empirical Result Analysis

4.1. Descriptive Statistics

Table 2 shows the descriptive statistics of the variables. The mean value of audit fees (LNFEE) was 13.64 with a standard deviation of 0.57. The average of ETR, Ratediff, MaETR, MaRatediff, which are used to measure the degree of corporate tax avoidance (TA) is 0.23, 0.01, 0.21, 0.01 respectively, the standard deviation is 0.13, 0.10, 0.09, 0.09 respectively; the statistical values of control variables are within a reasonable range.

4.2. Regression Analysis

Table 3 presents the results of regression on tax avoidance and audit fees. The

coefficients of ETR and MaETR are −0.40 and −0.360 with significant level at 5%, respectively. At the same time, the coefficients of Ratediff, MaRatediff are 0.224, 0.270, respectively. From the results above, we deduce that tax avoidance will increase audit fees and verify the hypothesis 1.

There are differences between SOEs and non-SOEs in at least three aspects: motivation for tax avoidance, the degree of tax avoidance and the risk of bankruptcy, which may affect the relationship between tax avoidance and audit fees. To further explore the impact of property rights on the relationship of audit avoidance and audit fees, we divide the full sample into the sample of state-owned enterprises (SOEs) and non-state-owned enterprises (non-SOEs) according to property rights. As is shown in Table 4, the coefficient of tax avoidance

(ETR/Ratediff/MaETR/MaRatediff) is only significant in non-SOEs. The results show that non-SOEs are more tax-aggressive than SOEs, which results in higher audit fees. Hypothesis 2 is consistent with the result.

5. Robustness Test

As for the proxy variable of tax avoidance, we also refer to Donohoe (2014) to employ the tax-aggressive dummy-Avoid [14] . Because firms exhibiting low tax rates relative to their industry peers are more likely to be using aggressive tax strategies, if they are in the lowest quintile of ETR or MaETR by year within a two-digit SIC industry. Avoid equals 1 if a firm is in the lowest quintile of ETR or MaETR by year within an industry, and 0 otherwise. Regression results are show in Table 5 and come to be similar with above.

6. Conclusions

This paper selects the A-share non-financial listed companies in Shanghai and Shenzhen Stock Exchanges from 2012 to 2015 as a sample, explores the relationship between tax avoidance and audit fees, and introduces property rights of enterprises to further explore its impact on tax avoidance and audit fees. The study found that the audit fees and the degree of corporate tax avoidance are positively correlated. At the same time, compared to state-owned enterprises (SOEs), the greater the degree of avoidance in non-state-owned enterprises (non-SOEs), the higher the audit fees charged by accounting firms.

These findings may have important implications for auditors and regulators. For auditors, they may consider the impact of tax avoidance and pay more attention to enterprises that have high degree of tax avoidance to identify the audit risk brought by tax avoidance, which are beneficial to reduce audit risk effectively, especially when auditing non-state-owned enterprises (non-SOEs). For regulators, to improve the quality of transactions, restrain adverse selection behavior and achieve more efficient markets supervision, it is vital to reveal effective information of tax avoidance.

This paper explores the impact of property rights on relationship between tax avoidance and audit fees, the sample is only divided into state-owned enterprises (SOEs) and non-state-owned enterprises (non-SOEs). However, the sample can be divided into the central government-controlled enterprises, provincial government-controlled enterprises and non-state-owned enterprises for further research.