The Effect of Audit Experience on Audit Fees and Audit Quality

Abstract

Prior research on audit experience focuses on behavioral studies that are conducted by running experiments. Although these studies provide evidence on the role of experience in completing specific audit tasks, they do not shed light on how experience affects a complete audit engagement. We conduct an archival study to examine the effect of audit experience on audit fees and audit quality. Using unique data from China, where the signees of the audit report can be identified and linked with a government database containing personal information about certified public accountants, we find that experience is positively associated with audit fees and negatively associated with absolute discretionary accruals. Furthermore, we extend the research on personal characteristics of audit partners by considering the incremental effects of gender, education, engagement tenure, industry specialization, and client importance after controlling for overall audit experience. Overall, our results suggest that the auditors’ personal characteristics may serve as a signal of the level of care that will be exercised during the audit process. Our results also have implications for China’s recently announced regulation that would require localization of Big 4 offices in China.

Keywords

audit experience individual auditors audit fees audit quality audit partners

Introduction

We examine the effect of an individual’s overall audit experience on audit fees and audit quality using unique archival data. Prior behavioral research examines the relationship between individual auditors’ audit experience and audit judgments (e.g., Bonner, 1990; Libby & Frederick, 1990; Moeckel, 1990; Simnett, 1996). These behavioral studies argue that experienced auditors have greater knowledge and more developed memory structures than inexperienced auditors, which leads to more accurate judgments by experienced auditors, but the results of these studies are mixed. Bonner (1990) contends that the mixed findings could result from a failure to incorporate task-specific knowledge in the test instrument, suggesting that the results may be sensitive to the experimental design. Moreover, behavioral research focuses on manipulated specific audit tasks rather than the body of audit work completed for a client. To complement this stream of behavioral research, we examine the effects of audit experience using an archival approach.

Archival research based on individual auditors’ personal characteristics is extremely rare because of limited data availability. The difficulty of using individual auditors’ data in archival research usually arises for two reasons: (a) the audit partner’s name is not disclosed in the audit report, and (b) even where the partner’s name is disclosed in the audit report, information about the personal characteristics of that partner has not been publicly available. A few studies focus on audit partner tenure (e.g., Carey & Simnett, 2006; C. Y. Chen, Lin, & Lin, 2008; Manry, Mock, & Turner, 2008). These studies use the number of years for which a partner has served as the engagement partner for the same company. However, strictly speaking, audit partner tenure for a particular company is not a personal characteristic of the partner; rather, it is an engagement characteristic. In this study, we focus on the audit partner’s overall experience, that is, the number of years that an individual has been engaged in audit work, using unique data from China.

The Chinese audit context is unique because regulations require that the names of the two Certified Public Accountants (CPAs) who sign the audit report be disclosed and because it is possible to obtain some personal information about individual auditors from the Chinese Securities Regulatory Commission (CSRC), which oversees the licensing of auditors who audit listed companies. Thus, we are able to examine the effect of an auditor’s overall audit experience (as opposed to engagement experience) on audit fees and audit quality after controlling for gender, education, industry experience, engagement tenure, and client importance. We document a positive association between audit fees and the aggregate audit experience of the two signee auditors, suggesting that clients pay fee premiums to have their audits led by more experienced CPAs. As the audit fee reflects the hours and billing rates of the entire audit team, it is unlikely that this result is merely reflecting the higher charge out rate of more experienced lead auditors. More important, we find that clients of more experienced CPAs have lower absolute discretionary accruals than clients of less experienced CPAs, suggesting that experienced auditors are better able to constrain earnings management. Combined, this suggests that more experienced lead auditors can conduct higher quality audits and that clients are willing to pay a premium for this quality. Overall, our results suggest that the lead auditors’ personal characteristics may proxy for the level of care that will be exercised during the audit process, and thus may help mitigate information problems in a setting where the client cannot fully observe the quality of the audit.

Our study complements the work of Gul, Wu, and Yang (2013) who also use data from individual auditors in China. They focus on estimating individual auditor fixed effects and find that including these fixed effects significantly improves the explanatory power of regression models that seek to explain differences in audit quality (e.g., abnormal accruals, small profits). However, they find that only a small portion of the individual auditor fixed effects can be explained by demographic characteristics related to the auditor. Rather than focus on individual auditor fixed effects, we focus on the average effect of auditor experience. That is, we are not interested in how the auditor experience-audit quality relation differs between individual auditors; instead, we are interested in whether auditor experience is related to audit quality at a general level. Furthermore, while Gul et al. (2013) consider auditor experience specifically, they define it based on the number of years of Big N experience. We define experience as overall auditor experience which is important because, as Gul et al. (2013) note, the Big N have relatively small market shares in China. Finally, we also consider how auditor experience affects the price of an audit while Gul et al. (2013) do not.

Our study contributes to the literature in two additional ways. First, our research complements the behavioral research on audit experience. The behavioral studies document mixed evidence on the relationship between audit experience and audit judgments. As these studies were conducted by designing specific audit tasks, they are unable to test the effect of audit experience on the complete audit engagement. Thus, unlike behavioral research, we are able to examine the implications of audit experience in a broader context. Second, our study extends a small but important stream of archival research that focuses on individual audit partners. Prior studies focus on engagement tenure rather than overall audit experience. The effects of a longer engagement tenure should mainly reflect the auditor’s client-specific knowledge. By controlling for engagement tenure as well as auditors’ industry experience, we are able to isolate the role of overall experience—a measure of general knowledge—from client-specific knowledge and industry knowledge. We find evidence that greater general knowledge can reduce earnings management and increase audit quality.

Our research has also practical importance. In May 2012, the Ministry of Finance in China regulated that foreign partners in a Big 4 Chinese office should be not more than 20% by 2017, and the chief partner of the office should be a Chinese citizen. This regulation was announced 1 day after the Securities and Exchange Commission (SEC) charged Deloitte’s Shanghai office for failing to produce audit documents for a Chinese client, a violation of the Sarbanes–Oxley Act that requires the Public Company Accounting Oversight Board (PCAOB) to inspect non-U.S. auditors who audit U.S.-listed firms (e.g., “Accounting in China,” 2012). Together, these events raise questions about the quality and experience of Chinese auditors. Our results suggest that audit quality increases with overall audit experience, which suggests that increasing Chinese auditor involvement in Big 4 offices in China is unlikely to diminish audit quality in the long run as Chinese auditors gain greater experience as a result of the new regulation. Whether audit quality will suffer in the short term is outside the scope of our study, but is an interesting issue for future research.

The rest of the article is organized as follows. “Institutional Background” section introduces the institutional background. “Literature Review” section reviews the literature. “Hypotheses” section develops the hypotheses. “Research Design” section designs the research. “Empirical Results” section reports the results. “Conclusion” section concludes.

Institutional Background

On December 25, 1995, the Chinese Ministry of Finance issued the first batch of independent audit standards since the People’s Republic of China (PRC) was established in 1949.¹ These standards became effective on January 1, 1996. No. 7 Independent Audit Standards—Audit Report regulated the practices of CPAs issuing audit reports. According to the standards, CPAs issuing audit reports were required to sign and stamp their individual names, and to also stamp their accounting firm’s name on the audit report. Nevertheless, there were no specific requirements on whose names or how many CPAs’ names should be signed and stamped on the audit report.

To reinforce the accountability of CPAs who issue audit reports, the Ministry of Finance released a supplementary regulation on the signing and stamping CPAs’ names on audit reports on July 2, 2001. This regulation, Notice of the Ministry of Finance on Related Issues of CPAs’ Signature and Seal on Audit Reports, required that the audit report should be signed and sealed by two CPAs who are licensed to issue the related audit report² and should be sealed by the accounting firm of the two CPAs. More specifically, the regulation requires that the signers should include the audit partner, chief CPA or an entrusted deputy chief CPA, who makes the final review of the audit, and another CPA who is in charge of the daily work on the audit.

On April 13, 2003, the Ministry of Finance issued a revised version of No. 7 Independent Audit Standards—Audit Report, which became effective on July 1, 2003. Similar to the original version, the revised version states that audit reports should be signed and sealed by CPAs, and also sealed by the accounting firm. In addition, the revised version provided a template for the audit reports, which indicated that two CPAs were required to sign and stamp their names.

On February 15, 2006, the Ministry of Finance issued No. 1501 Audit Standards of Chinese CPAs—Audit Report to replace No. 7 Independent Audit Standards—Audit Report. This batch of audit standards became effective from January 1, 2007. With respect to the requirements on the signature and seal of CPAs on audit reports, it is identical to the prior audit standard revised in 2003. The specific requirements on who should sign and seal audit reports are still the same as set out in the Ministry’s Notice that was issued in 2001. Thus, China has a relatively long history of requiring signatures of the two key auditors on an engagement.

Literature Review

Audit Experience

Prior research on audit experience was generally conducted by designing experiments or cases to examine the effect of audit experience on individual auditors’ judgments. Farmer, Rittenberg, and Trompeter (1987) document that experienced auditors are less likely to agree with the client’s preferred accounting treatment than inexperienced auditors, whereas Abdolmohammadi and Wright (1987) find a negative relation between audit experience and the likelihood of proposing an audit adjustment or qualified opinion. Biggs, Mock, and Watkins (1988) compare the analytical review of two audit managers with 7 years of audit experience and two senior auditors with 3 years of audit experience in a case study. They find that the senior auditors increased workload throughout the revenue cycle, while the audit managers did so more selectively. Libby and Frederick (1990) investigate the relation between error frequency knowledge and audit experience.³ Their findings indicate that perceptions about the source of the financial statement error are more accurate for experienced auditors than for inexperienced auditors.

Moeckel (1990) examines the effect of audit experience on the frequency of two types of memory errors including failure to integrate and reconstruction.⁴ She finds that inexperienced auditors fail to integrate more often than experienced auditors, whereas experienced auditors reconstruct more often than inexperienced auditors. As reconstructions impair experienced auditors’ ability to integrate and to detect the integration failures of their subordinates, Moeckel indicates that her findings challenge the intuition that experienced auditors will perform better at all aspects of work paper review. Bonner (1990) notes that the mixed results on the effect of audit experience on audit judgments may have arisen because some studies did not consider the role of task-specific knowledge in experimental designs.

In a study of the effect of information selection, information processing, and task complexity on predictive accuracy of auditors, Simnett (1996) finds that audit experience can mitigate some of the limitations arising from information selection, but not from information processing. Davis (1996) examines whether audit experience affects auditors’ approaches to selecting relevant information and whether selection of more relevant information improves auditors’ judgment performance for preliminary control risk assessments. By comparing new audit seniors and experienced audit seniors from a non-Big 6 accounting firm, he finds that experienced senior auditors exhibited a higher level of selective attention to relevant information than new senior auditors. Moreover, experienced senior auditors exhibited more consistency between the selected relevant information and the control risk assessment response, selected fewer cues, and made judgments in less time. In sum, these experimental studies provide valuable insights about the effects of experience in specific settings. We complement this research by examining the role of experience on the overall audit engagement.

Audit Fees and Audit Quality

Although there is a vast literature on audit fees, prior research has not considered how the personal characteristics of the lead auditors might affect the fee that is charged to clients. Instead, the prior research focuses on characteristics of the client firm, characteristics of the audit firm delivering the service, or characteristics of the specific engagement. In a meta-analysis, Hay, Knechel, and Wong (2006) find significant support for a Big 8/6/5/4 premium based on 85 different studies. Just as the Big N premium is due to reputational inferences made by investors, detailed knowledge of the personal characteristics of lead members of the audit team may influence the reputational inferences made by the client firm.

Audit Partner Tenure

Audit firm size and industry specialization are organization-level variables. Recently, several archival studies have focused on audit partner tenure, which is a partner/engagement-level variable. That is, it reflects the partner’s experience on a particular audit, but says nothing about the partner’s other clients or overall work experience. Still, these studies shed light on the value of partner rotation, which is an important issue from a regulatory perspective.

Carey and Simnett (2006) examine the relation between audit partner tenure and audit quality. They measure audit partner tenure as the number of years for which a partner has served as the engagement partner for the sample company. Using an Australian sample, they document that financial distressed firms with long audit partner tenure are less likely to receive a going-concern audit opinion. They also document some evidence that firms with long audit partner tenure are more likely to just beat earnings benchmarks, but no evidence on the association between audit partner tenure and discretionary working capital accruals. Using data from Taiwan, where audit reports of public companies should be certified by two audit partners whose names must be disclosed in the audit report, C. Y. Chen et al. (2008) identify audit partner tenure, and find that firms with long audit partner tenure have lower absolute discretionary accruals, inconsistent with the argument that long tenure audit partners provide lower quality services.

Manry et al. (2008) examine the association between audit partner tenure and discretionary accruals based on the data from multiple U.S. offices of three large international accounting firms. They find that audit partner tenure is negatively associated with discretionary accruals. Furthermore, they find that the significant and negative association between audit partner tenure and discretionary accruals is only for small clients with partner tenure of more than 7 years. Bedard and Johnstone (2010) investigate the relation between audit partner tenure and audit planning and pricing in United States. They find no association between audit partner tenure and planned audit hours, and a positive association between audit partner tenure and planned realization rates.⁵

Hypotheses

Behavioral auditing research suggests that audit experience affects the accuracy of auditors’ judgments. When auditors make judgments, they need to recall information necessary to perform tasks from memory (Bonner, 1990). The judgment accuracy depends on how much recalled information is matched to the task requirement (Libby & Frederick, 1990). As experienced auditors have developed vast and complex memory structures, they are able to use more adequate information in decision making, resulting in more accurate judgments. Moreover, these memory structures are precursors as they interpret the meaning and implications of domain-specific information, and thus affect auditors’ selection, understanding, and reaction to the task environment (Gibbins, 1984; Waller & Felix, 1984). Audit experience also affects the selection and weighting of information cues. Experienced auditors have knowledge structures that enable them to identify the particular information cues that should be selected and appropriately weighted to form their judgments (Bonner, 1990). Thus, audit experience can lead to more accurate audit judgments.

Signees of the audit report are audit partners or chief CPAs of accounting firms who are responsible for the final review of the audit, or CPAs who are in charge of the audit project. In a field experiment, Pratt and Jiambalvo (1981) find that in-charge auditors’ behaviors significantly affect the performance of audit teams supervised by those auditors. More experienced audit team leaders are more likely to behave in a way to increase audit team performance. These team leaders have more knowledge and expertise to effectively guide and monitor subordinates’ work.

The importance of personal characteristics can be linked to the economic theory of auditing. Auditing is a credence good because the client cannot determine the amount of service required ex ante and because the client cannot determine the quality of the service ex post (e.g., Causholli & Knechel, 2012; Emons, 1997). In such a setting, Dye (1993) expects that observable characteristics, such as the auditor’s wealth, will proxy for the unobserved audit quality. An auditor’s wealth is important because wealthy auditors have more to lose if they are sued for a substandard audit. DeAngelo (1981) makes a similar argument for audit firm size but focuses on quasi-rents. In our setting, the personal characteristics of the auditor may serve as an observable signal of the level of care that will be exercised during the audit process. Thus, knowledge of the lead auditors’ personal characteristics may mitigate information problems for the client firm.

Prior research (e.g., Francis, 1984; Palmrose, 1986) indicates that auditors charge higher fees on higher quality audits. If experienced auditors make better decisions as the experimental research suggests, they may increase audit fees to better reflect the higher quality of their audits. This quality differential may result in the higher audit fees charged by experienced signees. On the contrary, more experienced auditors may be able to conduct the audit more efficiently. However, the audit firm may not pass these cost savings on to clients.

Consequently, we examine two hypotheses. First, we examine whether there is a positive association between signees’ audit experience and audit fees. Second, to the extent experienced auditors charge higher fees, we consider whether the higher fees are due to experienced auditors providing higher audit quality. More formally, we test the following two hypotheses:

Hypothesis 1 (H1): Audit fees are positively associated with signees’ audit experience.

Hypothesis 2 (H2): Audit quality is positively associated with signees’ audit experience.

Research Design

Data Collection

We begin to select sample firms from the CSMAR-Audit Opinion database, which provides the data including audit fees of auditing the annual financial statements, the names of CPAs who signed the audit report, and the name of the audit firm, from all companies listed on both the Shanghai Stock Exchange and the Shenzhen Stock Exchange for the years 2007 to 2010.⁶ Using the data for years 1990 to 2010 from the CSMAR-Audit Opinion database, we identify the audit tenure of both signees and the audit firm.

We then manually collect the data on signees’ audit experience, gender, and education background from the official website of the CSRC (i.e., www.csrc.gov.cn), which is the government agency to regulate the Chinese securities markets.⁷ To be allowed to audit Chinese listed companies, audit firms and their CPAs must have a license from the CSRC. The CSRC provides the basic information of licensed audit firms and their individual licensed CPAs on its website including the number of years for which an individual CPA has been engaged in audit work, which is used in this study to measure the individual CPA’s audit experience.

Next, we manually collect the data on the number of each company’s business segments and the number of each company’s consolidated subsidiaries from the annual report, which is downloaded from the official websites of the Shanghai Stock Exchange and the Shenzhen Stock Exchange (i.e., www.sse.com.cn and www.szse.cn). We also acquire the data on each audit firm’s total revenue publicly released by the Chinese Institute of Certified Public Accountants (CICPA). Finally, we collect the data of financial statements from the Compustat Global database to compute other variables used in the statistical analysis. After the exclusion of firms with missing data and firms not covered by the databases, the final sample consists of 1,917 firm-year observations over the period 2007 to 2010.

Table 1 reports the breakdown of the final sample across 50 two-digit Standard Industrial Classification (SIC) industries, of which chemicals and allied products (18.36%), electrical and electronic equipment (10.12%), machinery and computer equipment (8.50%), primary metal industries (5.89%), transportation equipment (5.32%), and electric, gas, and sanitary services (5.32%) are the most representative industries in the sample.

Table 1.

Sample Breakdown by Industry.

Two-digit SIC code	Industry description	Frequency	%
12	Coal mining	25	1.30
16	Heavy construction contractors	28	1.46
20	Food and kindred products	75	3.91
22	Textile mill products	74	3.86
26	Paper and allied products	28	1.46
28	Chemicals and allied products	352	18.36
30	Rubber and miscellaneous plastics products	32	1.67
32	Stone, clay, glass, and concrete products	55	2.87
33	Primary metal industries	113	5.89
35	Machinery and computer equipment	163	8.50
36	Electrical and electronic equipment	194	10.12
37	Transportation equipment	102	5.32
38	Measuring and analyzing instruments	22	1.15
44	Water transportation	26	1.36
47	Transportation services	40	2.09
48	Communications	20	1.05
49	Electric, gas, and sanitary services	102	5.32
50	Wholesale trade-durable goods	40	2.09
51	Wholesale trade-non-durable goods	37	1.93
53	General merchandise stores	52	2.71
73	Business services	50	2.61
99	Nonclassifiable establishments	37	1.93
Others		250	13.04
Total		1,917	100.00

Note. SIC = Standard Industrial Classification.

Models

We test the first hypothesis by estimating the following regression model:

\begin{matrix} A F E E = a_{0} + a_{1} S E X P + a_{2} S F M + a_{3} S E D U + a_{4} S T E N + a_{5} S I S + a_{6} A F T E N + a_{7} A F I S \\ + a_{8} A F S I Z E + a_{9} A F I N T + a_{10} S I Z E + a_{11} R O A + a_{12} D E B T + a_{13} L O S S + a_{14} N S E G \\ + a_{15} N S U B S + a_{16} R E C + a_{17} I N V + a_{18} A U O P + i n d u s t r y d u m m i e s + ε, \end{matrix}

where AFEE = audit fees, measured as the natural logarithm of audit fees for auditing the company’s annual financial statements; SEXP = signee audit experience, measured as the natural logarithm of the sum of years for which two signees have engaged in audit work; SFM = female signees, coded “2,”“1,” and “0” if both signees, one of the two signees, and no signees are female, respectively; SEDU = signee education, measured as the sum of two signees’ education level, which is coded “3,”“2,” and “1” if a signee has a PhD degree, master’s degree, and undergraduate degree or others, respectively; STEN = signee tenure, measured as the natural logarithm of the sum of years for which the two signees have audited the client; SIS = signee industry specialization, measured as the sum of the two signees’ industry specialization, which is the ratio of the sales of the clients of a signee in a two-digit SIC industry to the total sum of the sales of all companies in that industry; AFTEN = audit firm tenure, measured as the natural logarithm of the maximum number of years for which the audit firm has audited the client; AFIS = audit firm industry specialization, measured as the ratio of the sales of the clients of the audit firm in a two-digit SIC industry to the total sum of the sales of all companies in that industry; AFSIZE = audit firm size, measured as the natural logarithm of the total revenue of the audit firm; AFINT = audit firm international alliance, coded “1” if the audit firm is a member of an international audit firm alliance, and “0” otherwise; SIZE = client size, measured as the natural logarithm of total assets; ROA = return on assets, measured as the ratio of income before extraordinary items to total assets; DEBT = debt ratio, measured as the ratio of long-term debt to total assets; LOSS = loss making, coded “1” if income before extraordinary items is negative and “0” otherwise; NSEG = segment, measured as the natural logarithm of the number of business segments; NSUBS = subsidiary, measured as the natural logarithm of the number of consolidated subsidiaries; REC = receivables intensiveness, measured as the ratio of receivables to total assets; INV = inventory intensiveness, measured as the ratio of inventory to total assets; AUOP = audit opinion, coded “1” if a modified audit opinion is issued, and “0” otherwise.

In Equation 1, the coefficient on SEXP is expected to be positive and significant if the first hypothesis is supported.⁸ Because we use the panel data for years 2007 to 2010, standard errors are clustered by year.

We include control variables for other personal characteristics that might affect audit fees or quality. Numerous studies examine the role of gender in business settings. For example, Adams and Ferreira (2009) find that gender-diverse boards devote more effort to monitoring, and prior research (e.g., Lund, 2008; Nguyen, Basuray, Smith, Kopka, & McGulloh, 2008) suggests that women may be more ethical and effective than men. In an audit context, a few experimental studies examine the effect of gender on audit judgments. Estes and Reames (1988) find that female auditors differ from male auditors in terms of confidence in their materiality decisions. J. Chung and Monroe (2001) find that women auditors outperform male auditor in an inventory valuation task. Gold, Hunton, and Gomaa (2009) find that female auditors are less likely to be swayed by unverified client-provided explanations, but are more (less) likely to be persuaded by male (female) clients when compared with male auditors. To control for gender effects, we include SFM in the model.

As education level reflects an individual’s knowledge and skill base (Hambrick & Mason, 1984), we control for SEDU in the model. Very few studies examine the effects of education on auditor’s decision making. Estes and Reames (1988) find no difference in auditors’ material decisions based on various measures of education. Grant, Ciccotello, and Dickie (2002) find the 150-hr rule adopted in many U.S. states had only marginal impact on student’s CPA exam success. In a survey, Ethridge and Heminway (1993) find only 31% of respondents thought individuals with an accounting master’s degree perform better than individuals with an accounting bachelor’s degree.

We include STEN in the model because, as discussed above, prior research (e.g., Carey & Simnett, 2006; Ghosh & Moon, 2005; Manry et al., 2008) indicates a positive effect of audit partner tenure and audit firm tenure on audit quality. We control for STEN so we can disentangle the effects of the audit partner’s firm-specific knowledge and general knowledge.

To examine the effects of industry-specific knowledge, we control for industry specialization for each signees of the audit report. Solomon, Shields, and Whittington (1999) find that industry-specific knowledge leads to more accurate non-error frequency knowledge. Owhoso, Messier, and Lynch (2002) find that industry specialists are better at detecting errors, and Low (2004) finds that specialists provide superior risk assessment and audit planning decisions. Thus, SIS and STEN allow us to examine the effect of general audit experience that is incremental to the lead auditors’ industry-specific knowledge and firm-specific knowledge.

Consistent with the prior literature, we also include several variables measured at the audit firm level. Palmrose (1986) documents a positive relationship between audit fees and audit firm absolute size, and Becker, DeFond, Jiambalvo, and Subramanyam (1998) find that auditor size is positively associated with audit quality. Thus, AFSIZE is added to the model. As audit firms develop areas of expertise and can build teams of experts (e.g., Cahan, Jeter, & Naiker, 2011), we include AFIS to capture industry specialization at the audit firm level. As prior research shows an association between audit firm tenure and earnings quality (e.g., Ghosh & Moon, 2005; Myers, Myers, & Omer, 2003), we include AFTEN. Finally, as audit firms may benefit from being part of an international accounting firm alliance (e.g., Carson, 2009), we include AFINT. As China is a developing economy, such alliances may be particularly important in terms of users’ perceptions about audit quality and reputation.

In addition, we include several client-related variables. SIZE, NSUBS, and REC in the model as prior research (e.g., Abbott, Parker, Peters, & Raghunandan, 2003; Francis & Simon, 1987; Simon & Francis, 1988) finds that audit fees are positively associated with the three variables. As Carcello, Hermanson, Neal, and Riley (2002) indicate that LOSS, SEG, and INV positively affect audit fees, these three variables are also added to Equation 1. According to Larcker and Richardson (2004), ROA and DEBT are included in the model. AUOP is used as a control variable in Equation 1 as it is found to be positively related to audit fees (Francis & Simon, 1987; Simon & Francis, 1988). Finally, industry dummies are added in the model to control for fixed industry effects.

The second hypothesis is tested by estimating the following regression model:

\begin{matrix} A D A C = b_{0} + b_{1} S E X P + b_{2} S F M + b_{3} S E D U + b_{4} S T E N + b_{5} S I S + b_{6} S C I M + b_{7} A F T E N \\ + b_{8} A F I S + b_{9} A F C I M + b_{10} A F S I Z E + b_{11} A F I N T + b_{12} S I Z E + b_{13} R O A + b_{14} D E B T \\ + b_{15} L O S S + b_{16} S G R O W + b_{17} C F O + i n d u s t r y d u m m i e s + ε . \end{matrix}

All the variables in Equation 2 have been defined for Equation 1 except for the following variables: ADAC = absolute discretionary accruals, measured as the absolute value of discretionary accruals; SCIM = signee client importance, measured as the sum of the two signees’ client importance, which is the ratio of the natural logarithm of the client’s total assets to the sum of the natural logarithm of total assets of all clients audited by a signee; AFCIM = audit firm client importance, measured as the ratio of the natural logarithm of the client’s total assets to the sum of the natural logarithm of total assets of all clients audited by the audit firm; SGROW = sales growth, measured as the annual percentage change in sales; CFO = cash flow, measured as the absolute value of cash flow from operations deflated by total assets.

Like prior research (e.g., Becker et al., 1998; Francis, Maydew, & Sparks, 1999), we use discretionary accruals to measure audit quality. Following S. Chen, Sun, and Wu (2010), we also use the propensity of issuing modified audit opinions as an alternative measure of audit quality. To measure discretionary accruals, we estimate the Jones (1991) model within each two-digit SIC industry based on the data of all Chinese firms from the Compustat Global database in each year:

\frac{A C C}{T A_{- 1}} = c_{0} \frac{1}{T A_{- 1}} + c_{1} \frac{Δ S A L E S}{T A_{- 1}} + c_{2} \frac{P P E}{T A_{- 1}} + ε,

where ACC = total accruals, measured as income before extraordinary items minus cash flows from operation; TA₋₁ = total assets at the beginning of the year; ΔSALES = annual change in sales; PPE = gross property, plant, and equipment.

After estimating Equation 3 within each two-digit SIC industry in each year, we compute the residual value for each sample firm to measure discretionary accruals.⁹

In Equation 2, we expect a negative and significant coefficient on SEXP if the second hypothesis is supported. As discussed for Equation 1, signees’ other characteristics such as SFM, SEDU, STEN, and SIS, and the audit firm’s characteristics such as AFTEN, AFIS, AFSIZE, and AFINT are also included in Equation 2. In addition, SCIM and AFCIM are included in the model because S. Chen et al. (2010) find that auditor client importance affects audit quality. As Armstrong, Barth, Jagolinzer, and Riedl (2010) suggest that larger firms have higher earnings quality, SIZE is added in Equation 2. Kothari, Leone, and Wasley (2005) indicate that earnings performance affects the measurement of discretionary accruals. Thus, we include ROA in the model. We control for DEBT because financial leverage may have the twofold effect on corporate governance (Klein, 2002; Jensen & Meckling, 1976). Francis, LaFond, Olsson, and Schipper (2004) indicate that accrual quality is lower for loss-making firms. Thus, we add LOSS in the model. As high sales growth could be associated with lower audit quality (Cahan & Zhang, 2006), SGROW is added in the model. Like H. Chung and Kallapur (2003), we add CFO to control for the effect of cash flow from operations on the measurement of discretionary accruals.

We also test the second hypothesis by estimating the following regression:

\begin{matrix} A R A = d_{0} + d_{1} S E X P + d_{2} S F M + d_{3} S E D U + d_{4} S T E N + d_{5} S I S + d_{6} S C I M + d_{7} A F T E N \\ + d_{8} A F I S + d_{9} A F C I M + d_{10} A F S I Z E + d_{11} A F I N T + i n d u s t r y d u m m i e s + ε, \end{matrix}

where ARA is audit reporting aggressiveness, measured as the difference between predicted probability from the following logistic model and AUOP (Gul et al., 2013):

A U O P = e_{0} + e_{1} Q U C I K + e_{2} A R I N V + e_{3} O T H R E C + e_{4} R O A + e_{5} S I Z E + e_{6} D E B T + e_{7} L O S S + ε,

where QUICK = quick ratio, measured by the sum of cash and short-term investments, notes receivables, and accounts receivables divided by current liabilities; ARINV = accounts receivables and inventory, measured by the sum of accounts receivables and inventory divided by total assets; OTHREC = other receivables, measured by other receivables divided by total assets.

Higher values of ARA indicate auditors’ lower propensity to issue modified audit opinions, and thus higher audit reporting aggressiveness. If signees with high audit experience more effectively audit financial statements, their clients are more likely to receive modified audit opinions. We expect a negative and significant coefficient on SEXP in Equation 4 if the second hypothesis is supported.

Empirical Results

All variables are defined in the appendix. The descriptive statistics of variables are reported in Table 2. The mean and median natural logarithm values of audit fees are 13.219 and 13.122 in RMB yuans, equivalent to RMB ¥550,730 and RMB ¥499,818 for the mean and median audit fees, respectively. The mean and median natural logarithm values of audit experience are 3.042 and 3.219 years, indicating that the mean and median of the sum of years for which two signees have engaged in audit work are 21 and 25 years. Table 3 presents the Pearson correlations of independent variables. The correlation between SIS and AFIS is extremely high (r = .84). To mitigate this multicollinearity issue, SIS is dropped from Equations 1, 2, and 4. We find that the results are not substantially changed if SIS is not dropped from the models.

Table 2.

Descriptive Statistics.

Variable	M	Median	SD	Q1	Q3
AFEE	13.219	13.122	0.585	12.821	13.527
ADAC	0.063	0.044	0.062	0.019	0.086
ARA	−0.001	0.003	0.164	0.000	0.018
SEXP	3.042	3.219	0.650	2.833	3.466
SFM	0.547	0.000	0.647	0.000	1.000
SEDU	2.252	2.000	0.501	2.000	2.000
STEN	1.547	1.609	0.556	1.099	1.946
SIS	0.108	0.027	0.241	0.010	0.074
SCIM	1.176	1.192	0.503	0.728	1.498
AFTEN	1.307	1.386	0.811	0.693	2.079
AFIS	0.087	0.036	0.141	0.013	0.091
AFCIM	0.034	0.023	0.041	0.013	0.038
AFSIZE	9.827	9.887	0.993	9.073	10.354
AFINT	0.364	0.000	0.481	0.000	1.000
SIZE	7.780	7.664	1.156	7.015	8.474
ROA	0.032	0.029	0.064	0.010	0.058
DEBT	0.063	0.016	0.098	0.000	0.087
LOSS	0.114	0.000	0.318	0.000	0.000
NSEG	0.751	0.693	0.627	0.000	1.099
NSUBS	1.838	1.792	0.860	1.386	2.398
REC	0.136	0.116	0.102	0.054	0.190
INV	0.157	0.136	0.123	0.067	0.214
SGROW	0.172	0.138	0.314	−0.007	0.303
CFO	0.058	0.053	0.075	0.015	0.096

Note. All variables are defined in the appendix.

Table 3.

Pearson Correlations.

Variable	SFM	SEDU	STEN	SIS	SCIM	AFTEN	AFIS	AFCIM	AFSIZE	AFINT	SIZE	ROA	DEBT	LOSS	AUOP	NSEG	NSUBS	REC	INV	SGROW	CFO
SEXP	.09***	−.10***	.14***	.02	−.02	.11***	.04	.03	−.15***	−.01	.08***	.06**	.01	−.06**	−.05**	.02	.07***	−.00	−.02	.05**	−.03
SFM		−.11***	−.00	.07***	.04*	.04	.04	.01	−.01	.06**	.02	.01	.01	−.02	−.03	.01	.05**	−.02	.03	.01	.01
SEDU			−.05**	−.02	.00	−.05**	−.04	.07***	.06**	.08***	−.02	−.03	−.03	−.04	.04	.00	−.04*	.01	−.03	.04*	.02
STEN				.02	−.18***	.56***	.01	−.01	−.01	.03	.12***	.02	.11***	−.04*	−.13***	.12***	.05**	−.15***	.03	−.01	.09***
SIS					.01	.03	.84***	−.03	.09***	.09***	.31***	.05**	.03	−.07***	−.04*	.11***	.19***	.02	−.02	.08***	.06**
SCIM						−.09***	.02	.25***	−.02	.02	.09***	.05**	.02	−.06**	.03	.04	.03	−.01	−.03	.03	−.01
AFTEN							.01	.05*	−.02	.00	.14***	.03	.07***	−.06**	−.08***	.09***	.03	−.15***	−.04*	.01	.06**
AFIS								−.14***	.13***	.11***	.31***	.02	.08***	−.05*	−.07***	.08***	.18***	.01	−.04*	.09***	.03
AFCIM									−.15***	−.02	.03	−.01	.01	−.04	.02	.01	.02	.00	.04	.02	.01
AFSIZE										.37***	.12***	.02	.00	−.00	−.03	−.01	.06**	−.03	.03	.01	.06**
AFINT											.08***	.00	.02	−.00	.01	.00	.01	.03	.01	−.01	−.05*
SIZE												.13***	.38***	−.17***	−.24***	.05**	.26***	−.17***	.02	.22***	.06**
ROA													−.08***	−.68***	−.30***	−.04*	−.02	.06**	−.01	.30***	.34***
DEBT														.00	−.10***	.03	.02	−.32***	−.21***	.07***	−.02
LOSS															.31***	.01	−.01	−.08***	−.04*	−.24***	−.16***
AUOP																.01	.02	.05**	−.10***	−.18***	−.11***
NSEG																	.33***	−.07***	−.00	−.03	−.04*
NSUBS																		.00	.04	.01	.00
REC																			.13***	.03	−.16***
INV																				.03	−.18***
SGROW																					.09***

Note. All variables are defined in the appendix.

***

, **, and * denote significance at 1%, 5%, and 10% level, respectively (two-tailed tests).

Table 4 provides the results on audit fees. We find that the coefficient on SEXP is positive and significant (t statistic = 3.86), consistent with H1. Thus, clients pay audit fee premiums to signees with higher audit experience, suggesting that experienced CPAs conduct higher quality audits. We also find that the coefficient on SEDU is positive and significant (t statistic = 4.13), suggesting that signees with graduate degrees receive audit fee premiums. Audit firms with high industry specialization charge high audit fees (t statistic = 14.88). Audit firm size positively affects audit fees (t statistic = 2.60). In addition, we find significant evidence that audit fees are positively associated with the client size, return on assets, loss making, the number of subsidiaries, receivables intensiveness, and issuance of modified audit opinions, but are negatively associated with inventory intensiveness.

Table 4.

Results for Audit Fees.

Variable	Predicted sign	Coefficient	t statistic
Intercept		9.447	32.13***
SEXP	+	0.069	3.86**
SFM		0.023	1.16
SEDU		0.036	4.13**
STEN		−0.014	−0.76
AFTEN		0.016	1.51
AFIS		0.375	14.88***
AFSIZE		0.084	2.60*
AFINT		0.004	0.57
SIZE		0.301	33.54***
ROA		0.470	3.34**
DEBT		−0.102	−1.98
LOSS		0.074	4.37**
NSEG		−0.015	−1.35
NSUBS		0.147	14.64***
REC		0.227	3.21**
INV		−0.225	−3.47**
AUOP		0.262	5.96***
Industry dummies			Included
N			1,917
F statistic			19.71**
Adjusted R²			55.55%

Note. All variables are defined in the appendix.

***

and * denote significance at 1% and 10% level, respectively (two-tailed test unless sign predicted).

Table 5 reports the results on absolute discretionary accruals. We find that the coefficient on SEXP is negative and significant (t statistic = −6.44), which supports H2. Thus, experienced signees are more effective in oversight of clients’ financial reporting. This shows that those CPAs can provide higher quality services to clients, and suggests that the fee premium associated with SEXP reported in Table 4 is due to the higher quality product being delivered by experienced lead auditors.

Table 5.

Results for Absolute Discretionary Accruals.

Variable	Predicted sign	Coefficient	t statistic
Intercept		0.119	2.75*
SEXP	−	−0.004	−6.44***
SFM		−0.000	−1.95
SEDU		0.001	0.21
STEN		−0.005	−1.91
SCIM		−0.003	−1.40
AFTEN		−0.001	−0.52
AFIS		−0.008	−0.44
AFCIM		0.064	1.46
AFSIZE		0.001	0.76
AFINT		−0.004	−2.10
SIZE		−0.008	−2.84*
ROA		0.098	1.38
DEBT		−0.022	−2.45*
LOSS		0.043	4.91**
SGROW		0.015	2.54*
CFO		0.032	0.42
Industry dummies			Included
n			1,600
F statistic			74.85***
Adjusted R²			9.61%

Note. All variables are defined in the appendix.

***

, **, and * denote significance at 1%, 5%, and 10% level, respectively (two-tailed test unless sign predicted).

Table 6 presents the results on audit reporting aggressiveness. We find an insignificant coefficient on SEXP (t statistic = 0.43), inconsistent with our conjecture that clients of experienced signees are more likely to receive modified audit opinions. We also find a positive and significant coefficient on SFM (t statistic = 2.57), indicating that female audit partners have lower propensity to issue modified audit opinions. Nevertheless, we suggest that these results should be cautiously interpreted because the percentage of firms receiving modified audit opinions in our sample is only 5.68% and the F statistic is not significant, which may reflect the inappropriateness of this analysis.

Table 6.

Results for Audit Reporting Aggressiveness.

Variable	Predicted sign	Coefficient	t statistic
Intercept		−0.113	−1.75
SEXP	−	0.004	0.43
SFM		0.007	2.57*
SEDU		0.001	0.12
STEN		0.014	1.39
SCIM		−0.011	−1.74
AFTEN		0.006	0.91
AFIS		−0.028	−1.22
AFCIM		0.028	0.24
AFSIZE		0.007	1.49
AFINT		−0.013	−1.63
Industry dummies			Included
n			1,600
F statistic			3.93
Adjusted R²			3.45%

Note. All variables are defined in the appendix.

***

, **, and * denote significance at 1%, 5%, and 10% level, respectively (two-tailed test unless sign predicted).

We conduct additional analyses as follows. First, we test the hypotheses by controlling for experienced signees’ self-selection bias. Based on the auditor selection model used in prior research (e.g., Khurana & Raman, 2004), we estimate the following logistic model:

S E X P I = f_{0} + f_{1} S H O R T + f_{2} L O N G + f_{3} P E + f_{4} I S S U E + f_{5} S I Z E + f_{6} D E B T + f_{7} L O S S + ε,

where SEXPI = an indicator variable of experienced signees; SHORT = absolute value of short-term accruals in income deflated by sales; LONG = absolute value of long-term accruals in income deflated by sales; PE = ratio of price to earnings per share; ISSUE = new equity issue, coded “1” if change in equity is greater than 10% and “0” otherwise.

The indicator variable is coded “1” if the sum of the first and second signees’ audit experience is greater than 32 years, which is the 75th percentile of the distribution of the two signees’ audit experience. We use the fitted value from Equation 6 to compute the inverse Mills ratio. Then, we include the inverse Mills ratio in Equations 1, 2, and 4 to correct for the self-selection bias (Heckman, 1979).

Table 7 includes the results after controlling for the self-selection bias. From Equation 1, we find that the audit experience is positively associated with audit fees (t statistic = 3.37). From Equation 2, we find that audit experience is negatively associated with absolute discretionary accruals (t statistic = −3.12). Thus, the results after allowing for the experienced signees’ self-selection bias still support H1 and H2. From Equation 4, we find that there is an insignificant association between audit experience and the issuance of modified audit opinions (t statistic = 0.64), which is consistent with the findings without controlling for the experienced signees’ self-selection bias.

Table 7.

Results After Controlling for Self-Selection Bias.

Variable	Audit fees			Absolute discretionary accruals				Audit reporting aggressiveness
	Predicted sign	Coefficient	t statistic	Variable	Predicted sign	Coefficient	t statistic	Variable	Predicted sign	Coefficient	t statistic
Intercept		9.337	27.03***	Intercept		0.110	2.20	Intercept		−0.115	−1.81
SEXP	+	0.067	3.37**	SEXP	−	−0.004	−3.12**	SEXP	−	0.007	0.64
SFM		0.013	0.72	SFM		−0.001	−1.07	SFM		0.006	1.67
SEDU		0.031	1.58	SEDU		−0.001	−0.37	SEDU		0.005	0.50
STEN		−0.010	−0.63	STEN		−0.004	−1.77	STEN		0.019	1.87
AFTEN		0.019	1.77	SCIM		−0.003	−1.01	SCIM		−0.022	−4.62**
AFIS		0.402	14.86***	AFTEN		−0.001	−0.30	AFTEN		−0.002	−0.40
AFSIZE		0.093	2.52*	AFIS		−0.013	−0.56	AFIS		−0.017	−0.94
AFINT		−0.010	−0.88	AFCIM		0.064	2.39*	AFCIM		0.049	0.56
SIZE		0.305	26.09***	AFSIZE		0.002	0.88	AFSIZE		0.005	1.14
ROA		0.368	2.64*	AFINT		−0.004	−1.68	AFINT		−0.003	−0.65
DEBT		−0.126	−1.57	SIZE		−0.006	−1.67	IMR		−0.001	−0.26
LOSS		0.097	4.05**	ROA		0.046	0.59
NSEG		−0.015	−0.97	DEBT		−0.032	−2.56*
NSUBS		0.153	10.03***	LOSS		0.041	4.01**
REC		0.214	2.68*	SGROW		0.013	1.51
INV		−0.193	−2.34	CFO		0.080	1.30
AUOP		0.227	3.14*	IMR		−0.001	−2.70*
IMR		0.011	1.91
Industry dummies			Included	Industry dummies			Included	Industry dummies			Included
n			1,732	n			1,451	n			1,451
F statistic			20.91**	F statistic			21.21**	F statistic			4.28
Adjusted R²			56.19%	Adjusted R²			9.62%	Adjusted R²			4.77%

Note. All variables are defined in the appendix.

***

, **, and * denote significance at 1%, 5%, and 10% level, respectively (two-tailed test unless sign predicted).

Second, we estimate Equation 2 by using income-increasing or income-decreasing discretionary accruals as the dependent variable separately. Like the results on absolute discretionary accruals, we find that the coefficient on SEXP remains negative and significant for income-increasing discretionary accruals (non-tabulated t statistic = −3.35). The coefficient on SEXP is positive and insignificant for income-decreasing discretionary accruals (non-tabulated t statistic = 0.81). Overall, our results still support H2 when signed discretionary accruals are used as the dependent variable in Equation 2; in particular, more experienced auditors are more likely to constrain aggressive income-increasing accruals.

Third, we separate SEXP into SEXP1 and SEXP2 in the equations, where SEXP1 is the first signee’s audit experience and SEXP2 is the second signee’s audit experience.¹⁰ From Equation 1, we find that the coefficients on SEXP1 and SEXP2 are positive and significant (non-tabulated t statistics = 3.59 and 2.92, respectively), consistent with H1. From Equation 2, we find that the coefficients on SEXP1 and SEXP2 are negative and significant (non-tabulated t statistics = −3.34 and −4.21, respectively), consistent with H2. From Equation 4, we find negative and insignificant coefficients on SEXP1 and SEXP2 (non-tabulated t statistics = −0.43 and 0.29, respectively). Thus, the results are inconsistent with the results on SEXP.

Fourth, we re-run the analyses by measuring SEXP as the maximum value of the first or second signees’ audit experience. Similar to the results on the sum of the first and second signees’ audit experience, the maximum value of SEXP1 or SEXP2 is positively associated with audit fees (non-tabulated t statistic = 2.93), and is negatively associated with absolute discretionary accruals (non-tabulated t statistic = −3.37). Likewise, clients’ likelihood of receiving modified audit opinions is not associated with the maximum value of SEXP1 or SEXP2 (non-tabulated t statistic = 0.83).

Fifth, we measure discretionary accruals using the performance-matched discretionary accruals developed by Kothari et al. (2005). For absolute discretionary accruals, we find a negative but insignificant coefficient on SEXP (non-tabulated t statistic = −0.70). The coefficient on SEXP is negative and significant for income-increasing discretionary accruals (non-tabulated t statistic = −4.65), and is insignificant for income-decreasing discretionary accruals (non-tabulated t statistic = 0.13), consistent with the results of non-performance-matched discretionary accruals. The coefficient on SEXP1 is negative and significant (non-tabulated t statistic = −1.71), whereas the coefficient on SEXP2 is insignificant (non-tabulated t statistic = −0.00). Overall, we still find some evidence to support H2. Recently, Keung and Shih (2014) show that the performance matching approach significantly underestimates discretionary accruals and increases the frequency of Type II errors.

Sixth, in the spirit of Teoh and Wong (1993), we examine whether audit experience affects the association between stock returns and accounting earnings by running the following regression:

\begin{matrix} R E T = g_{0} + g_{1} Δ E + g_{2} S E X P + g_{3} Δ E \times S E X P + g_{4} S F M + g_{5} Δ E \times S F M + g_{6} S E D U + g_{7} Δ E \times S E D U \\ + g_{8} S T E N + g_{9} Δ E \times S T E N + g_{10} S I S + g_{11} Δ E \times S I S + g_{12} S C I M + g_{13} Δ E \times S C I M + g_{14} A F T E N \\ + g_{15} Δ E \times A F T E N + g_{16} A F I S + g_{17} Δ E \times A F I S + g_{18} A F C I M + g_{19} Δ E \times A F C I M + g_{20} A F S I Z E \\ + g_{21} Δ E \times A F S I Z E + g_{22} A F I N T + g_{23} Δ E \times A F I N T + i n d u s t r y d u m m i e s + ε . \end{matrix}

In Equation 6, RET is annual stock returns. ΔE is measured as the annual change in income before extraordinary items deflated by the beginning of year market value of common equity. If investors view unexpected earnings of firms with more experienced auditors as being more credible, the coefficient g₃ (i.e., the earnings response coefficient) will be positive and significant.

We find an insignificant coefficient on ΔE×SEXP (non-tabulated t statistic = −1.51), suggesting that investors either do not recognize or appreciate the higher quality audit of experienced individual auditors. However, we caution that such a test is a joint test of audit quality and market efficiency. Thus, an alternative explanation for the insignificant coefficient is that the Chinese stock market is inefficient because of its speculative nature and susceptibility to rumors and investor sentiment. Groenewold, Wu, Tang, and Fan (2004) and others provide evidence of inefficiency, although the overall evidence is mixed. Chong, Lam, and Yan (2012) find that while the efficiency of the Chinese stock market has improved over time, some trading rules remain profitable in the latest period they examine (2009-2010).

Conclusion

We conduct an archival study to examine the effect of audit experience on audit fees and audit quality. Using the unique data from China, we find that the overall audit experience of two signees of the audit report is positively associated with audit fees, and is negatively associated with absolute discretionary accruals after controlling for other personal characteristics of the lead auditors, audit firm characteristics, and client characteristics. Our findings suggest that more experienced CPAs provide higher quality audit services than less experienced CPAs.

Prior research on audit experience focuses on behavioral studies that are conducted by running experiments. These studies only provide experimental and task-based results. Based on those results, it is not possible to determine how audit experience affects the complete audit engagement. Thus, archival evidence could complement the prior behavioral studies in an important way, but obtaining personal characteristics of auditors in a large sample has been difficult. Using unique data from China, we are able to help fill this void in the literature. Furthermore, we extend the research on audit partners by considering the incremental effects of gender, education, engagement tenure, industry specialization, and client importance after controlling for overall audit experience. Overall, our results suggest that the lead auditors’ personal characteristics may serve as a signal of the level of care that will be exercised during the audit process. In this way, personal characteristics help mitigate adverse selection problems for the client in a setting where audit quality is not fully observable, that is, a credence good.

Our study complements the work of Gul et al. (2013). Although they examine the effects of demographic characteristics of Chinese auditors at an individual level, we focus on the average effect across individual auditors. Finally, our study has implications for the recently announced Chinese regulation that would require the localization of Big 4 offices in China. Our results suggest that audit quality increases with the experience of Chinese auditors. Thus, giving Chinese auditors more experience by increasing their involvement in Big 4 offices will benefit auditing in China in the long run.

Footnotes

Appendix

Acknowledgements

We thank Bharat Sarath (the editor), Matthew Reidenbach, two anonymous reviewers, and participants at the 2012 American Accounting Association (AAA) Annual Meeting for their valuable comments. We appreciate Huanwen Jiao, Siwen Li, Hong Zhao, and Ya Zhou for their research assistance.

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

Notes

References

Abbott

L. J.

Parker

Peters

Raghunandan

(2003). The association between audit committee characteristics and audit fees. Auditing: A Journal of Practice & Theory, 22, 17-32.

Abdolmohammadi

Wright

(1987). An examination of the effects of experience and task complexity and audit judgments. The Accounting Review, 62, 1-13.

Accounting in China: Internal controls. (2012, May19). The Economist, p. 74.

Adams

Ferreira

(2009). Women in the boardroom and their impact on governance and performance. Journal of Financial Economics, 94, 291-309.

Armstrong

C. S.

Barth

M. E.

Jagolinzer

A. D.

Riedl

E. J.

(2010). Market reaction to the adoption of IFRS in Europe. The Accounting Review, 85, 31-61.

Becker

DeFond

Jiambalvo

Subramanyam

(1998). The effect of audit quality on earnings management. Contemporary Accounting Research, 16, 1-24.

Bedard

J. C.

Johnstone

K. M.

(2010). Audit partner tenure and audit planning and pricing. Auditing: A Journal of Practice & Theory, 29, 45-70.

Biggs

S. F.

Mock

T. J.

Watkins

P. R.

(1988). Auditors use of analytical review in audit program design. The Accounting Review, 63, 148-161.

Bonner

S. E.

(1990). Experience effects in auditing: The role of task-specific knowledge. The Accounting Review, 66, 72-92.

10.

Cahan

S. F.

Jeter

D. C.

Naiker

(2011). Are all industry specialists the same?Auditing: A Journal of Theory & Practice, 30, 191-222.

11.

Cahan

S. F.

Zhang

(2006). After Enron: Audit conservatism and ex-Andersen clients. The Accounting Review, 81, 49-82.

12.

Carcello

J. V.

Hermanson

D. R.

Neal

T. L.

Riley

R. A., Jr.

(2002). Board characteristics and audit fees. Contemporary Accounting Research, 19, 365-384.

13.

Carey

Simnett

(2006). Audit partner tenure and audit quality. The Accounting Review, 81, 653-676.

14.

Carson

(2009). Industry specialization by global audit firms networks. The Accounting Review, 84, 355-382.

15.

Causholli

Knechel

W. R.

(2012). An examination of the credence attributes of an audit. Accounting Horizons, 26, 631-656.

16.

Chen

C. Y.

Lin

C. J.

Lin

Y. C.

(2008). Audit partner tenure, audit firm tenure, and discretionary accruals: Does long auditor tenure impair earnings quality?Contemporary Accounting Research, 25, 415-445.

17.

Chen

Sun

S. Y. J.

(2010). Client importance, institutional improvements, and audit quality in China: An office and individual auditor level analysis. The Accounting Review, 85, 127-158.

18.

Chong

T. T.-L.

Lam

T.-H.

Yan

I. K.-M.

(2012). Is the Chinese stock market really inefficient?China Economic Review, 23, 122-137.

19.

Chung

Kallapur

(2003). Client importance, nonaudit services, and abnormal accruals. The Accounting Review, 78, 931-955.

20.

Chung

Monroe

(2001). A research note on the effects of gender and task complexity on an audit judgment. Behavioral Research in Accounting, 13, 111-125.

21.

Davis

(1996). Experience and auditor’s selection of relevant information for preliminary control risk assessments. Auditing: A Journal of Practice & Theory, 15, 16-37.

22.

DeAngelo

(1981). Auditor size and auditor quality. Journal of Accounting & Economics, 1, 183-199.

23.

Dye

R. A.

(1993). Auditing standards, legal liability, and auditor wealth. Journal of Political Economy, 101, 887-914.

24.

Emons

(1997). Credence goods and fraudulent experts. RAND Journal of Economics, 28, 107-119.

25.

Estes

Reames

D. D.

(1988). Effects of personal characteristics on materiality decisions: A multivariate analysis. Accounting & Business Research, 18, 291-296.

26.

Ethridge

J. R.

Heminway

J. R.

(1993). Industry’s view of the 150-hour requirements. The CPA Journal, 63, 67-69.

27.

Farmer

T. A.

Rittenberg

L. E.

Trompeter

G. M.

(1987). An investigation of the impact of economic and organizational factors on audit independence. Auditing: A Journal of Practice & Theory, 7, 1-14.

28.

Francis

J. R.

(1984). The effect of audit firm size on audit prices: A study of the Australian market. Journal of Accounting & Economics, 6, 133-151.

29.

Francis

J. R.

LaFond

Olsson

P. M.

Schipper

(2004). Costs of equity and earnings attributes. The Accounting Review, 79, 967-1010.

30.

Francis

J. R.

Maydew

Sparks

(1999). The role of Big 6 auditors in the credible reporting of accruals. Auditing: A Journal of Practice & Theory, 18, 17-34.

31.

Francis

J. R.

Simon

D. T.

(1987). A test of audit pricing in the small client segment of the U.S. audit market. The Accounting Review, 62, 145-157.

32.

Ghosh

Moon

(2005). Auditor tenure and perceptions of audit quality. The Accounting Review, 80, 585-612.

33.

Gibbins

(1984). Propositions about the psychology of professional judgment in public accounting. Journal of Accounting Research, 22, 103-125.

34.

Gold

Hunton

Gomaa

(2009). The impact of client and auditor gender on auditors’ judgments. Accounting Horizons, 23, 1-18.

35.

Grant

C. T.

Ciccotello

C. S.

Dickie

(2002). Barriers to professional entry: How effective is the 150-hour rule?Journal of Accounting and Public Policy, 21, 71-93.

36.

Groenewold

Tang

S. H. K.

Fan

X. M.

(2004). The Chinese stock market: Efficiency, predictability and profitability. Cheltenham, UK: Edward Elgar.

37.

Gul

F. A.

Yang

(2013). Do individual auditors affect audit quality? Evidence from archival data. The Accounting Review, 88, 1993-2023.

38.

Hambrick

D. C.

Mason

P. A.

(1984). Upper echelons: The organization as a reflection of its top managers. Academy of Management Review, 9, 193-206.

39.

Hay

Knechel

Wong

(2006). Audit fees: A meta-analysis of the effect of demand and supply attributes. Contemporary Accounting Research, 23, 141-191.

40.

Heckman

(1979). Sample selection bias as a specification error. Econometrica, 7, 53-61.

41.

Jensen

M. C.

Meckling

(1976). Theory of the firm: Managerial behaviour, agency costs, and ownership structure. Journal of Financial Economics, 3, 305-360.

42.

Jones

J. J.

(1991). Earnings management during import relief investigation. Journal of Accounting Research, 29, 193-228.

43.

Keung

Shih

(2014). Measuring discretionary accruals: Are ROA-matched models better than the original Jones-type models?Review of Accounting Studies, 19, 736-768.

44.

Khurana

I. K.

Raman

K. K.

(2004). Litigation risk and the financial reporting credibility of Big 4 versus non-Big 4 audits: Evidence from Anglo-American countries. The Accounting Review, 79, 473-498.

45.

Klein

(2002). Audit committee, board of director characteristics, and earnings management. Journal of Accounting & Economics, 33, 375-400.

46.

Kothari

S. P.

Leone

A. J.

Wasley

C. E.

(2005). Performance matched discretionary accrual measures. Journal of Accounting & Economics, 39, 163-197.

47.

Larcker

Richardson

(2004). Fees paid to audit firms, accrual choices, and corporate governance. Journal of Accounting Research, 42, 625-658.

48.

Libby

Frederick

D. M.

(1990). Experience and the ability to explain audit findings. Journal of Accounting Research, 28, 348-367.

49.

Low

K. Y.

(2004). The effects of industry specialization on audit risk assessments and audit-planning decisions. The Accounting Review, 79, 201-219.

50.

Lund

D. B.

(2008). Gender differences in ethics judgment of marketing professionals in the United States. Journal of Business Ethics, 77, 501-515.

51.

Manry

Mock

Turner

(2008). Does increased partner tenure reduce audit quality?Journal of Accounting, Auditing & Finance, 23, 553-572.

52.

Moeckel

(1990). The effect of experience on auditors’ memory errors. Journal of Accounting Research, 28, 368-387.

53.

Myers

J. N.

Myers

L. A.

Omer

T. C.

(2003). Exploring the term of the auditor-client relationship and the quality of earnings: A case for mandatory auditor rotation?The Accounting Review, 78, 779-799.

54.

Nguyen

T. N.

Basuray

M. T.

Smith

W. P.

Kopka

McGulloh

(2008). Moral issues and gender differences in ethical judgment using Reidenbach and Robin’s (1990) multidimensional ethics scale: Implications in teaching of business ethics. Journal of Business Ethics, 77, 417-430.

55.

Owhoso

V. E.

Messier

W. F.

Lynch

J. G.

(2002). Error detection by industry specialized teams during sequential audit review. Journal of Accounting Research, 40, 883-900.

56.

Palmrose

Z.-V.

(1986). Audit fees and auditor size: Further evidence. Journal of Accounting Research, 24, 97-110.

57.

Pratt

Jiambalvo

(1981). Relationships between leader behaviors and audit team performance. Accounting, Organization & Society, 6, 133-142.

58.

Simnett

(1996). The effect of information selection, information processing and task complexity on predictive accuracy of auditors. Accounting, Organizations & Society, 21, 699-717.

59.

Simon

D. T.

Francis

(1988). The effects of auditor change on audit fees: Tests of price cutting and price recovery. The Accounting Review, 63, 255-269.

60.

Solomon

Shields

M. D.

Whittington

O. R.

(1999). What do industry-specialist auditors know?Journal of Accounting Research, 37, 191-208.

61.

Teoh

S. H.

Wong

T. J.

(1993). Perceived auditor quality and the earnings response coefficient. The Accounting Review, 68, 346-366.

62.

Waller

W. S.

Felix

W. I.

(1984). The auditor and learning from experience: Some conjectures. Accounting, Organizations & Society, 9, 383-406.