審計(jì)人員的激勵(lì)、認(rèn)知和行為

2002年薩班斯-奧克斯利法案(SOX)限制審計(jì)師為同一客戶提供審計(jì)和非審計(jì)服務(wù),因?yàn)橛纱水a(chǎn)生的經(jīng)濟(jì)聯(lián)系將損害審計(jì)獨(dú)立性和破壞了審計(jì)質(zhì)量(美國(guó)眾議院,2002)。然而,一些研究認(rèn)為某些特定的非審計(jì)服務(wù)其實(shí)可以給審計(jì)人員提供必要的知識(shí)來(lái)理解客戶的業(yè)務(wù)和事務(wù)操作(Copeland, 2002; Goldwasser, 2002; Pletz, 2002; Shedlarz, 2002)。通過(guò)豐富知識(shí),從而提高審計(jì)效率(Simunic, 1984; Whisenant et al., 2003; Kinney et al., 2004; Flaming, 2005; Antle et al., 2006; Gaynor et al., 2006)。因此, 對(duì)于審計(jì)獨(dú)立性和審計(jì)質(zhì)量是優(yōu)先提倡經(jīng)濟(jì)激勵(lì)措施的凈影響還是非審計(jì)服務(wù)的豐富知識(shí)出現(xiàn)了長(zhǎng)期的爭(zhēng)論。
 
Chapter 3 Auditors’ Incentives, Perceptions, and Behaviours
 
The Sarbanes-Oxley Act of 2002 (SOX) restricts auditors on offering both audit and non-audit services to one client under the claim that the resulting economic bonding will jeopardise auditor independence and undermine the audit quality (U.S. House of Representatives, 2002). However, some academic researches state that specific non-audit services can offer auditors with necessary knowledge to understand the clients’ business and transaction operations (Copeland, 2002; Goldwasser, 2002; Pletz, 2002; Shedlarz, 2002) and thus improve audit effectiveness via knowledge spill-overs (Simunic, 1984; Whisenant et al., 2003; Kinney et al., 2004; Flaming, 2005; Antle et al., 2006; Gaynor et al., 2006). Therefore, a longstanding debate towards the net effects of economic incentives and knowledge spill-over of non-audit services on auditor independence and audit quality has been emerged. Furthermore, whether there is an association between audit and non-audit fees are strongly debated as the link between such two services can indicate the net effects of economic bonds and knowledge spill-over and test whether audit services are influenced by non-audit services provided by incumbent auditors. In addition, auditor independence can also be evaluated by behaviours of auditors’ - propensity to issue going-concern opinions, which is under the assumption that auditors with jeopardised auditor independence are not motivated to issue a going-concern opinion. These three issues will be discussed as follows.
 
3.1經(jīng)濟(jì)激勵(lì)和知識(shí)溢出Economic incentives versus knowledge spill-over
Well before the scandal of Enron and Arthur Andersen, Bazerman et al. (1997) indicated that the auditing system had been corrupted by the economic incentives which induce auditors to please their clients and compromise their independence. This is not because auditors are more corruptible than other but human nature is easily induced by economic incentives (Bazerman and Moore, 2011). By acknowledged the fact, it is worthy further investigation about whether the benefits of knowledge spill-over can outweigh such impairment to audit quality. Therefore, effects of economic bond on auditor independence and effects of knowledge spill-over on audit quality will be discussed, followed by different views of prior literatures on the relationship between audit and non-audit fees. 
 
3.1.1經(jīng)濟(jì)債券對(duì)審計(jì)獨(dú)立性的影響Effects of economic bond on auditor independence
Auditors have an economic motivation to provide non-audit services to their clients for the reason that non-audit services are generally considered more profitable. Although it is presumed by regulatory rules that non-audit services provided by incumbent auditors will jeopardise auditor independence and threaten audit quality, documented evidence about the possibility that auditor independence is jeopardised resulting from the economic bonding derived from the provisions of non-audit services is mixed (e.g., DeAngelo, 1981; Dopuch and King, 1991; Wines, 1994; Firth, 1997; Frankel et al., 2002; Gaynor, 2002; Ashbaugh et al., 2003; Kinney et al., 2004; Bedard et al., 2008; Kanagaretnam et al., 2011; Krishnan and Yu, 2011; Knechel and Sharma, 2012; Schmidt, 2012). For example, Dopuch and King (1991) argue that prohibition of auditors from providing both audit and non-audit services does not help to improve efficiencies but might trigger auditors to choose non-audit services over audit services from the experiment. However, Asare et al. (2005) propose a contradict opinion that client acceptance or staffing decisions of the auditor are not influenced by their potential decisions to maintain non-audit service engagements.
 
3.1.2知識(shí)溢出對(duì)審計(jì)質(zhì)量的影響Effects of knowledge spill-over on audit quality
Non-audit services might help auditors gain more knowledge of clients thus the probability of discovering fraud and errors might increase. It is suggested that maintaining a given degree of auditor independence, non-audit services are considered to increase audit quality (Lennox, 1999), which is the knowledge spill-over effects of non-audit services. Evidence from Knechel and Sharma (2012) also support this argument as the clients companies purchasing a relative high level of non-audit services prior to the Sarbanes-Oxley Act of 2002 (SOX) are more likely to involve large increase in audit lags after the announcement of the Sarbanes-Oxley Act of 2002 (SOX), which indicates the knowledge spill-over have positive impact on audit effectiveness.
 
Following the initial research about the definition of knowledge spill-over is from Simunic (1984) and Palmrose’s (1986) presumption that knowledge gained by providing non-audit services to clients might flow to the audit services thus produce audit efficiencies, tremendous researches have investigated the association between audit fees and non-audit fees for the purpose of examining the assertion that pricing effects of knowledge spill-over (economics of scope) can be explained by associations between audit and non-audit fees, which will be explained in details as follows.
 
3.1.3審計(jì)和非審計(jì)費(fèi)用之間的聯(lián)系A(chǔ)ssociation between audit and non-audit fees
Previous researches have three different views towards the association between audit fees and non-audit fees. It seems that earlier studies probe a significantly positive relationship and no relationship between such two services fees and the current studies are more likely to support a negative association instead. These three viewpoints will be discussed respectively as follows.
 
A number of earlier studies support the notion that a significantly positive association between audit fees and non-audit fees (Simunic, 1984; Simon, 1985; Palmrose, 1986, Turpen, 1990, Barkess and Simnett, 1994; Firth, 2002) exists. However, by investigating data of audit hours offered by a large accounting profession, Davis et al. (1993) state that although the association between audit effort and non-audit fees is positive, such relationship is weak with no significant evidence of knowledge flows between non-audit and audit feed. Antle et al. (2006) further Davis et al.’s (1993) study by applying system of simultaneous equations rather than single-equation models, which might cause biased estimates and inferences, and find similar positive associations between audit and non-audit fees, which supports the knowledge spill-over or economics of scope of joint provisions of audit and non-audit services. More detailed research carried out by Joe and Vandervelde (2007), who related knowledge transfer to audit risk assessments in order to provide more direct evidence, suggest that audit risk assessment process can be benefited from non-audit services provided by incumbent auditors. This is demonstrated by the fact that auditors who performed joint provisions of audit and non-audit services are more likely to assess a higher audit risk than those who only offered audit services but were accessible to the relevant working papers of non-audit services.
 
On the other hand, some researches indicate that no association exists between audit fees and non-audit fees (Abdel-khalik, 1990; Raghunandan et al., 2003). Raghunandan et al. (2003) also adopt system of simultaneous equations, but reach the opposite conclusions as Antle et al. (2006) do, find that the joint determined effects between audit and non-audit fees but no economics of scope of performing both audit and non-audit services. In addition, Bedard et al. (2008) suggest that no sufficient evidence to support the presumption that the joint provision of audit and non-audit services tend to jeopardise auditor independence.
 
Apart from above debate, current researches argue that the association between audit and non-audit fees is significantly negative. For example, Krishnan and Yu (2011) document that knowledge spill-over is a two-way process as it can flow from audit part to non-audit part as well as non-audit part to audit part by adopting a two-stage squares model and applying in a large sample. They also suggest that certain characteristics of auditors and clients, such as auditors in Big 4 firms, long audit tenure, highly dependent clients, extent of tax services, and client importance, can trigger significant knowledge spill-over. Similar conclusion is drawn from Krishnan and Visvanathan (2011) that insights from offering tax services can contribute to audit quality in the condition that both audit and tax services are offered by same audit firm, which is consistent with knowledge spill-over theory. Recent literatures show that abnormal non-audit fees are positively connected with abnormal loan loss provisions of small banks (Kanagaretnam et al., 2010) and with more negative results of auditor class action litigation (Schmidt, 2012), which suggest that non-audit services are negatively considered by some stakeholders. In addition, by investigation the difference in audit report lags pre- and post- the enact of Sarbanes-Oxley Act of 2000 (SOX), Knechel and Sharma (2012) discover that the audit firms involving a high level of non-audit services fees before the enact of Sarbanes-Oxley Act of 2000 (SOX) have a relative increase in audit report lags in post-SOX period, providing a consistent presumption of knowledge spill-overs benefits that non-audit services provided by incumbent auditors are beneficial to clients without lower the level of audit effectiveness.#p#分頁(yè)標(biāo)題#e#
 
To summarise, the inconsistency in the research of the association between audit and non-audit fees might result from a variety of reasons. Firstly, the selection of non-audit services types is likely to cause the different results as earlier studies are more likely focus on management advisory services (MAS), which is banned by the Sarbanes-Oxley Act of 2002 (SOX) now, and current studies emphasis more on non-audit tax service (NATS), which is still allowed by SOX despite more restrictions in its scope than before. In addition, SOX requires prior approval from audit committees on non-audit services (see Chapter 4), which might trigger different purchase behaviours of the client companies than prior Sarbanes-Oxley Act of 2000 (SOX) thus leading to different research results. It is indicated that audit fee has constituted a larger part of total fees paid to auditors (Krishnan and Yu, 2011) and the effects of Sarbanes-Oxley Act of 2000 (SOX) on the dynamics of audit and non-audit services should be considered in further research.
 
3.2 持續(xù)經(jīng)營(yíng)的意見(jiàn)傾向問(wèn)題Propensity to issue going-concern opinions
Apart from examining the relationship between audit and non-audit service fees, archival studies, which focus on auditors’ behaviours, adopt the proxy of auditors’ propensity to issue going-concern opinion to investigate whether joint provisions of audit and non-audit services will impair auditor independence.
 
3.2.1審計(jì)師的獨(dú)立性和持續(xù)經(jīng)營(yíng)意見(jiàn)Auditor independence and going-concern opinions
The audit reports play an important role in signalling users about the financial statements of forthcoming going-concern dilemmas thus it is a significant tool for auditors to communicate with financial statements users about their findings. Considering the importance of auditors’ opinions, auditors are required to perform in an objective manner and free from pressure from clients to issue a clean opinion. Therefore, it is assumed that as audit independence is widely acknowledged as the possibility that the auditor will report a breach which was discovered in the financial reports (Watts and Zimmerman, 1983), thus auditors who are with compromised independence have a low probability to issue going-concern opinions in the condition that such opinions are warranted (DeFond et al., 2002). Because auditor independence is closely related to the issue of audit failure, which includes situations which auditors do not issue going-concern opinions to companies which will be filed for liquidity subsequently (Blacconiere and DeFond, 1997). Therefore, legislators have focused critical attention on situations in which companies have filed for liquidity without receiving a going-concern opinion in advance (Geiger and Raghunandan, 2002). Consequently, current researches adopt auditors’ propensity to issue going-concern opinions as an appropriate indicator to evaluate the influences of non-audit services fees on auditor independence and audit quality.
 
There is one obvious advantage of using the auditor’s propensity to issue going-concern opinions as the indicator to assess auditor independence as it is more direct and less noisy (DeFond et al., 2002) than management earnings (e.g., discretionary accruals and tendency to meet earnings targets) do as the impact of auditors on earnings of clients features seems more indirect as well as empirical troubles exist in discretionary measurement (Dechow et al., 1995; Guay et al., 1996; McNicholes, 2000; Collins and Hribar, 2002) (see Chapter 5.1).
 
3.2.2非審計(jì)服務(wù)和持續(xù)經(jīng)營(yíng)意見(jiàn)Non-audit services and going-concern opinion
However, prior researches have mixed opinions towards the relationship of going-concern opinions issuances and auditor-provided non-audit services fees. A majority of prior literatures indicate that there is no relationship between the level of non-audit services fees and the auditor’s propensity to issue going-concern opinions (Barkess and Simnett, 1994; Craswell, 1999; Lennox, 1999; DeFond et al., 2002; Geiger and Rama, 2003; Callaghan et al., 2009). Other literatures suggest that the relationship exists between non-audit services fees and issuances of qualified audit reports or going concern opinions is negative (Wines, 1994; Sharma, 2001; Sharma and Sidhu, 2001; Firth, 2002; Lindberg and Beck, 2004; Venuti, 2004; Basioudis et al., 2008; Robinson, 2008; Carson et al., 2011).
 
By investigating top 500 listed Australian companies from 1986 to 1990, Barkess and Simnett (1994) find no correlation identified between the degree of non-audit services provided by incumbent auditors and the type of audit reports they issued, which supports the assertion that auditor independence seems not to be jeopardised by joint provision of non-audit services. Using the public information from Australian listed companies, Craswell (1999) also reaches the same conclusion by case studies and questionnaires. At the same period, Lennox (1999) examines the British financially distressed companies, which are in a high possibility to receive a going-concern opinion (Callaghan et al., 2009), and concludes that the association between non-audit service fees and audit qualifications is insignificant. Further investigation is conducted in U.S. and after the mandated disclosure of non-audit fees. It is suggest that no relationship exists between non-audit services fees and the incumbent auditors’ propensity to issue a going concern opinion because related reputation losses and litigation exposure, which are market-based institutional incentives, can strengthen auditor independence and outweigh the economic incentives derived from higher fees (DeFond et al., 2002). In addition, Geiger and Rama (2003) reach same conclusion by comparing financially distressed companies which received going-concern audit opinions with those not receiving going-concern opinions. Callaghan et al. (2009) incorporate unexpected fees and client importance to their research to strengthen the validity of the conclusion thus it is robust to various sensitivity analyses.
 
On the other hand, Wines (1994) find that because of relatively high remuneration of auditors which derived from high level of non-audit service fees, auditors are reluctant to qualify financial statements of a company if they also offer non-audit services to this company. Firth (2002) finds a negative link between non-audit services fees and issuances of qualified reports by analysing information of listed companies in U.K. In addition, by examining audit reports of financially distressed companies in the U.K., Basioudis et al. (2008) conclude that an important relationship between the magnitudes of non-audit services fees with the issuance of a going-concern modified audit opinion, supporting the contention that a high level purchase of non-audit services fees have an significant impact on going-concern reporting judgments for financial distressed companies in the U.K. Carson et al. (2011) state that non-audit services provided by incumbent auditors have negative influences on the propensity of auditors to issue a going-concern opinion.
 
Apart from these inconsistent findings, other factor, such as auditor industry specialist, might have influences on auditors’ propensity to issue going-concern opinions. For example, Lim and Tan (2008) suggest that auditors from a higher level of industry specialist have a higher probability to issue a going-concern report than those with non-specialist because industry specialist auditors are concerned more about their reputation losses and litigation exposure as well as benefit more from knowledge spill-over from the joint provision of non-audit services than those non-specialists do. Therefore, audit quality measured by propensity to issue going-concern opinion is positively increasing with non-audit services provided by industry specialist auditors compared with non-specialist auditors.
 
3.2.3導(dǎo)致不同結(jié)果的可能原因Possible reasons for inconsistent findings
The reasons for mixed results are worthy further investigation. Callaghan et al. (2009) argue that the jurisdictions of data used in researches and economic situations of samples are two main factors causing the different outcomes. In addition, the validity of data gathered in earlier studies is not sufficient, particularly in the U.S. because non-audit services are not prohibited and enterprises have no responsibility to disclose non-audit fees. In contrast, although non-audit services are not prohibited in the U.K. and Australia, the fees of such services are required to be disclosed to public (Lennox, 1999).
 
第四章客戶的動(dòng)機(jī)、認(rèn)知和行為Chapter 4 Clients’ Incentives, Perceptions, and Behaviours
 
For the perspective of management of client firms, auditor independence and audit quality have been assessed by various proxies, including stock and bond markets’ reactions to non-audit services disclosure (Abbott et al., 2011). In this part, capital market’s reactions are classified as cost of debt and equity capital and earnings response efficiencies. In addition, pre-approval of audit committees and shareholders’ voting against auditor ratification are also from the perspectives of the clients’ companies, which intends to strengthen the monitoring process of purchases of non-audit services and maintain due auditor independence.
 
4.1資本市場(chǎng)的反應(yīng)Capital market reaction
Prior studies suggest that firms with a relative high level of non-audit services purchases are involved in higher costs of debt and equity capital (Brandon et al., 2004; Dhaliwal et al., 2008; Nam and Ronen, 2012) and lower earnings response coefficients (Krishnan et al., 2005; Francis and Ke, 2006; Higgs and Skantz, 2006), which will be discussed respectively.
 
4.4.1債務(wù)和股權(quán)資本的成本Cost of debt and equity capital
Auditor independence is important to establishing public confidence in capital markets (Kanagaretnam et al., 2010) thus investors’ perceptions towards auditor independence will influence their investments decisions, which might affect the cost of debt and equity capital of clients’ companies.
 
It is argued that non-audit services fees are supposed to undermine auditor independence, reduce the reliability of financial statements, and thus increase the firm’s cost of capital, including cost of debt and cost of equity. However, prior research focuses more on impact of non-audit services fees on debt market than equity markets for the following reasons: First, bondholders gain scheduled interests and not tend to seek high risk investment as equity-holders do (Brandon et al., 2004), thus they rely more on public financial statements and audited accounting information (Holthausen and Watts, 2001); Second, the bond market accepts more explicit intervention than the equity market do and the bond pricing models are better identified and less subject to the misspecification of the asset pricing model (Mansi et al., 2004).
 
Instead exploring the relationship between levels of non-audit services purchases with cost of debt capital directly, Brandon et al. (2004) investigated the association between the magnitude of non-audit services provided by incumbent auditors and the client’s bond ratings. The empirical results indicate that the amount of non-audit services purchases is negatively related with the client’s bond rating. The reason for the evaluation of bond ratings is because the information bond ratings convey about the value of the client’s firm and its ability to repay its debt obligations results in fluctuations on both the stock and bond markets. According to Ziebart and Reiter (1992), bond ratings are negatively associated with cost of debt capital. However, Brandon et al.’s (2004) research only focuses on bond market without analysis on equity markets. Dhaliwal et al. (2008) also ascertains that non-audit services fees are positively associated with the cost of debt to client’s companies, however, this situation is only valid for investment-grade enterprises. For enterprises with non-investment-grade debt, non-audit services fees are negatively related to investors’ reliance on accounting information, such as return of assets (ROA).
 
Unlike only emphasis on debt market, Nam and Ronen (2012) investigated the relationship between non-audit services and cost of capital and the bid/ask spread. By analysing data from 2001 to 2003, Nam and Ronen (2012) suggest that non-audit services are negatively associated with the cost of capital and bid/ask spread, while positively related to the predictability and forecasting.
 
For identifying the reacts of markets to impacts of non-audit services, Abbott et al. (2013) separates the influences of non-audit services fees into two stages: pre-SOX and post-SOX, that is, nondisclosure environment and disclosure environment. It is argued that the financing activity has significantly affected the purchase behaviours in the disclosure environment, which indicate the mandated disclosure of non-audit services fees promote the transparency of financial information, and provide the investors to clear signals on fluctuations of costs of capital.
 
4.4.2股票價(jià)格和收益反應(yīng)系數(shù)Stock price and earnings response coefficient (ERC)
As investors cannot participate in the operations and observe the underlying true earnings of the investees’ firms, they are willing to depend on the reported financial information, which requires external auditors to certify the reliability of financial information in accordance with GAAP. Therefore, it is plausible that if the credibility of auditor’s work is perceived by investors as high, investors will respond strongly to reported surprise earnings. In other word, investors’ response to an earnings surprise will depend on the perceived credibility of the earnings report (Teoh and Wong, 1993). It is applicable that the relevance earnings of non-audit services by a firm is measured by earnings response coefficients (ERCs) (Dhaliwal et al., 2006), which are commonly considered as a proxy to determine the auditor independence-in-appearance for investors’ perceptions of earnings and audit quality (Ghosh et al., 2009).
 
Prior studies suggest that ERCs in high non-audit serices purchases firms are lower than those firms with a lower level of non-audit services purchase for the reason that investors’ perception of the reliability of financial information, such as financial statements and financial reports, is adversely influenced by the provision of non-audit services (Dhaliwal et al., 2006; Higgs and Skantz, 2006; Francis and Ke, 2006; Krisnan et al., 2005). Therefore, these researches are supportive to regulators policy, such as SEC, about mandated disclosure to non-audit services fees and restrictions on types of non-audit services provided by incumbent auditors for the purpose of reducing threats to auditor independence and audit quality of incumbent auditors.
 
Ghosh et al. (2009) further their studies about independence-in-appearance by comparing the relative importance of non-audit services fee ratio and client importance. According to a larger sample from year 2001 to 2006, it is suggested that the non-audit services fee ratio is insignificantly related to ERCs while client importance is negatively related to ERCs. In other words, there is a negative perception of investors to high levels of client importance not non-audit fee ratio.
 
However, Eilifsen and Knivsfla (2013) argue that the relationship of effects of non-audit services provision on audit quality and regulatory restrictions and mandated disclosure on non-audit services is not cause and effects but perform interactively. Their results show that for low-quality (small size) audit firms, the relationship between non-audit services and ERC is negatively related after the mandated disclosure enacted. However, this conclusion is inconsistent with Ghosh et al. (2009), who suggest that ERCs do not vary with non-audit services fee ratio in any size firm. This inconsistency might result from the samples in Eilifsen and Knivsfla (2013) are from Norwegian data, which is a high regulatory environment and regulations have more significant impact on investors.
 
In conclusion, researches focus on capital reaction to impact of non-audit services provision suggest the level of non-audit services purchase is negatively associated with cost of capital and earnings response coefficients. In this case, researchers advocate the regulatory prohibitions of certain types of non-audit services and mandated disclosure of non-audit services fees. However, further investigation needs to conduct to assess various regulatory environments to better determine the regulations’ influences on investors’ perceptions of non-audit services (Eilifsen and Knivsfla, 2013).
 
4.2代理成本Agency costs
According to Krishnan (2003), the application of accrual-based earnings over cash flows in financial statements offers the opportunities for management to communicate their individual and internal information to improve the capability of earnings to show the underlying economic price. Nevertheless, outsiders (e.g., shareholders, investors) cannot observe the earnings directly, thus agency costs are produced in the process of separation of ownership and management.
 
In order to solve agency problems, earlier studies suggest that managers are likely to propose to hire independent auditors (Jensen and Meckling, 1976). Supporting this conclusion, evidence from Watts and Zimmerman (1983) find that 84% of New York Stock Exchange (NYSE) companies have voluntarily generally recruited independent auditors in 1926, a long time before the mandatory disclosures. Therefore, not only theoretical but also empirical evidence support that management are performing their best benefits to engage independent auditors.
 
Earlier studies suggest that companies with greater agency costs have a higher possibility to hire a higher quality of audit services, thus have a lower non-audit fee ratio (Parkash and Venable, 1993; Firth, 1997). Specifically, Parkash and Venable (1993) argue that the less concentrated ownership has less motivation for direct monitoring, thus the demand for auditor-provided non-audit services is relatively high in order to reduce the agency costs. Firth (1997) also suggests that companies with poor performance are more likely to purchase external consulting services to improve its profitability. However, these researches are earlier than the mandated disclosure of non-audit services fees regulated by SOX, therefore, the conclusions only document the impact of agency costs on auditor fees without addressing the role of market reactions to mandated non-audit services disclosures in creating subsequent year agency effects (Abott et al., 2011).
 
4.2.1審計(jì)委員會(huì)預(yù)先批準(zhǔn)Pre-approval from audit committees
Enron and other accounting scandals illustrate the importance of increasing audit committee governance accountability particularly because it is relevant with guarantee of auditor independence and audit quality (Gaynor et al., 2006). For the purpose of promotion effective corporate governance and addressing potential threats to auditor independence, the Securities and Exchange Commission (SEC) and Congress (through the Sarbanes-Oxley Act of 2002, U.S. House of Representatives, 2002) passed in order to increase audit committee’s accountability to investors for audit and financial reporting quality, which enacts that audit committees have the responsibility to pre-approve non-audit services provided by incumbent auditors when all non-audit services fees are more than 5% of total fees paid to the incumbent auditor in the prior year (Abott et al., 2003). Therefore, it seems that audit committee’s propensity to pre-approve of non-audit services is significantly influenced by the mandated disclosure of non-audit services fees.
 
In pre-approving non-audit services, the audit committee should consider the trade-off between auditor independence or objectivity and audit quality should be prepared to justify its decision to those to whom it is accountable: the board, management, and private or public investors (Braiotta, 2004). Prior literatures on the impact of non-audit disclosure on the audit committees are mixed. For example, Healy and Palepu (2001) and Hodge et al. (2004) state that managers are tend to make decisions to escape from improved transparency of financial reporting, which supports that decisions of audit committees are influenced by whether such decisions are disclosed to the public. Specifically, Hodge et al. (2004) find that audit committees with a high level of professionalism are publicly connecting fees to auditor independence. In addition, Gaynor et al. (2006) argue that non-audit services disclosures exacerbate the legal motivations and concerns of audit committees. However, Abott et al. (2003) suggest that the audit committees have legal motivations to control purchasing levels of non-audit regardless of the disclosure environment. This conclusion might need further evidence as Abbott et al. (2011) point out that no investigation on non-audit services fees in subsequent year after SOX. Most recent literature from Abbott et al. (2011) suggest that a significantly increase in the magnitude of audit committee effectiveness after mandated disclosure environment.
 
4.2.2股東投票反對(duì)批準(zhǔn)審計(jì)人員Shareholders’ voting against ratification of the auditor
Shareholders’ voting on auditor ratification plays an important role in corporate governance, even unlike pre-approval of non-audit services from audit committees, it is not mandating for shareholders to vote on auditor selection. However, it is suggested that companies with shareholders’ voting on auditor ratification have a lower level of subsequent restatements in financial statements and abnormal accruals, which indicate that the process of shareholders’ voting on auditor selection will promote a higher audit quality (Dao et al., 2012).
 
The mandated disclosure of non-audit services fees are presumed to be useful for shareholders to make their investment decisions. The earliest research focus shareholders’ voting against ratification of the auditor is from Glezen and Miller (1985), who point out that shareholders’ voting against auditor ratification has no relationship with the non-audit fees ratio during 1979 to 1981. It is reported that the average proportion of votes against auditor against auditor ratification is 0.6% and no different in this proportion during three years. However, Raghunandan (2003) find that the proportion of shareholders voting against or abstaining from ratification of the external auditor is positively correlated with the level of non-audit fee ratio by investigating voting of shareholders of 172 Fortune 100 companies. It is concluded that disclosure of fees paid to auditors have impact on shareholders’ voting decisions. Even in companies with a large amount of non-audit fees, the average shareholder ratification rate was around 97%, which indicate that the majority of shareholders did not realise the impaired of auditor independence, even in the presence of relatively high level of non-audit fees.
 
Further investigation connects the audit committee with shareholders’ voting against auditor ratification. Using a more diverse sample, Raghunandan and Rama (2003) find that the proportion of votes against auditor ratification is lower in the condition that the audit committee is comprised of independent directors and at least one member equipped with accounting or finance expertise.
 
Similar with audit committees’ pre-approval of non-audit services to clients, shareholders’ voting on auditor ratification is also significantly influenced by relevant regulations. As the SEC (2003) updated the regulations about disclosure of more detailed information about the fees paid to auditors by requiring non-audit services fees to be categorised into audit-related, tax, and other fees, which are considered to be useful for investors to assess the level of auditor independence and voting on ratifying the auditor, further studies start to probe the association between shareholders’ voting to auditor ratification and type of non-audit services. Using data from a sample of 248 S&P 1500 companies, Mishra et al. (2005) discover that the percentage of shareholders’ voting against auditor ratification in 2003 is positively related with not only the tax fee ratio but also other fee ratio, but involving a negative association  with the audit-related fee ratio.