Contemporary Accounting Research Vol. 26 No. 2 (Summer 2009) pp. 393–402 © CAAA doi:10.1506/car.26.2.3 Discussion of “Mandatory Audit Partner Rotation, Audit Quality, and Market Perception: Evidence from Taiwan”* E. MICHAEL BAMBER, University of Georgia LINDA SMITH BAMBER, University of Georgia 1. Introduction Chi, Huang, Liao, and Xie (2009) examine the effect of mandatory audit partner rotation on audit quality and the market’s perception of audit quality, measured using abnormal accruals and earnings response coefficients (ERCs). The effect of audit partner rotation on audit quality is an important question. Audit partner rota- tion is costly for auditing firms, especially with the increased frequency of rotation mandated by the Sarbanes-Oxley Act. Moreover, as DeFond and Francis (2005, 26) note, “there is a greater need than ever for objective scientific evidence to guide public policy-making in auditing now that it is explicitly controlled by a govern- ment agency [the Public Company Accounting Oversight Board] rather than the active product of competitive market forces”. To date, there has been little empirical evidence on the costs or benefits of audit partner rotation because of lack of data; in North America, engagement partners are not publicly identified. Chi et al. (2009) capitalize on the fact that in Taiwan, audit partners are identified in audit reports, and the two leading Taiwan- ese stock exchanges effectively mandated audit partner rotation in 2004. The Tai- wan experience provides a unique opportunity to examine the effects of mandatory audit partner rotation. The authors compare clients required to change their audit partners in 2004 with (a) clients not yet subject to mandatory rotation, (b) clients in the mandatory rotation sample, but measured in the year before mandatory rotation, and (c) clients in prior years whose audit partners rotated voluntarily. Regulators and even associations of professional auditors believe that audit partner rotation enhances audit quality. 1 Regulators are requiring auditor rotation out of concern that long tenure may erode auditor independence and/or hinder the auditor’s ability to develop creative and innovative audit programs due to compla- cency or overfamiliarity (Carey and Simnett 2006). The idea is that a fresh perspective * Accepted by Michael Willenborg. An earlier version of this paper was presented at the 2005 Con- temporary Accounting Research Conference, generously supported by the Canadian Institute of Chartered Accountants , the Certified General Accountants of Ontario , the Certified Man- agement Accountants of Ontario , and the Institute of Chartered Accountants of Ontario . We thank Michael Willenborg (associate editor) for giving us the opportunity to discuss this paper. We have benefited from helpful comments by Orie Barron, Jeremy Griffin, and Isabel Yanyan Wang. 394 Contemporary Accounting Research CAR Vol. 26 No. 2 (Summer 2009) will lead to higher-quality audits. As Chi et al. (2009) recognize, however, auditor rotation has costs as well as benefits. Given the loss of client-specific knowledge that potentially impairs the effectiveness and quality of the audit, there is a real need for scientific evidence on the extent to which rotation yields the benefits regu- lators anticipate. Researchers have responded by producing a flurry of studies on the association between auditor rotation (or nonrotation in the form of auditor ten- ure) and various measures of audit quality. As explained in the next section, Chi et al. (2009) contribute to this research stream. In this discussion, we focus on four main points concerning the Chi et al. 2009 study and where we see the greatest potential for future research on audit partner rotation and audit quality. First, we discuss how Chi et al. fit into the context of the extant literature, and we compare the challenges arising in tests of audit partner rotation versus audit firm rotation. Second, we consider the research design in Chi et al. 2009. As they suggest, the Taiwanese data provide an opportunity to address a research question that cannot be similarly addressed in a North American context. But the data and research design carry their own set of problems, including issues of statistical power. Third, we consider Chi et al.’s measures of audit quality. Inter- preting their study as providing evidence on audit quality assumes not only that audit quality and earnings quality are interchangeable, but also that abnormal accruals and ERCs are good measures of earnings quality and, by extension, audit quality. While such arguments are common in the archival audit literature, we believe it is time to move beyond these generic proxies, especially if we want regu- lators to act on our research in settings as specific as audit partner rotation. Finally, our discussion concludes with suggestions for future research. 2. Size matters: Effects of rotating audit firms versus audit partners Auditor rotation can be accomplished by changing either the audit firm or the audit partner. Most of the evidence on the effects of auditor rotation on audit quality is based on audit firm rotation, although to date regulators have yet to require audit firm rotation. Instead, regulators in the United States, United Kingdom, Taiwan, Australia, and many other countries around the globe effectively require audit part- ner rotation. Because audit firm rotation is not mandatory, extant research explores the effect of voluntary audit firm (non)rotation, by documenting the relation between audit firm tenure and measures of audit quality. The general result from this litera- ture is that longer audit firm tenure is associated with higher-quality financial reporting. For example, Johnson, Khurana, and Reynolds (2002) and Myers, Myers, and Omer (2003) conclude that longer audit firm tenure constrains extreme absolute abnormal accruals. Mansi, Maxwell, and Miller (2004) find that longer audit firm tenure is associated with higher bond ratings and a lower cost of debt, and Ghosh and Moon (2005) find that longer audit firm tenure is associated with greater value-relevance of reported earnings and higher Standard & Poors’ ratings of the client’s shares. Using a field-based analysis, Bamber and Iyer (2007) find longer audit firm tenure mitigates acquiescence to the client’s preferences. 2 These results are consistent with critics’ arguments that the costs of audit firm rotation (in Discussion of “Mandatory Audit Partner Rotation and Audit Quality” 395 CAR Vol. 26 No. 2 (Summer 2009) terms of lost client-specific knowledge) exceed the benefits, thereby impairing rather than improving audit quality. We cannot extrapolate effects of audit firm rotation to audit partner rotation because the costs and benefits are likely quite different. In audit firm rotation, the new firm brings a new audit team, applies the (new) firm’s own audit methodology, and also applies additional new client procedures. In audit partner rotation, in most cases all that changes is one audit partner. What does not change is the audit firm and its audit methodology, prior working papers, and the firm’s history with the cli- ent, the rest of the audit team, and, in most cases, one of the two audit partners. Furthermore, Chen, Lin, and Lin (2008) note that mandatory partner rotation in Taiwan can be superficial in the sense that the partner rotates off a client only to rotate back on the following year, and maintains a relationship with the client over the entire period. Chen et al. (2008) report that over half the partners who rotated off in 2003 or 2004 rotated back onto the client after one year. Such superficial audit partner rotation is unlikely to have any effect on audit quality. 3 Given that changes resulting from audit partner rotation are much more limited in scope than changes from audit firm rotation, the costs and benefits are likewise more limited. Audit partner rotation can improve independence if problems arise from the individual partner and the new engagement partner significantly influences the rest of the audit team, but partner rotation cannot solve independence problems arising from audit firm culture. Audit partner rotation is less likely to compromise audit team competence because other elements of the audit team and technology remain in place, and legal liability increases incentives for the outgoing partner and the national office to train and monitor the incoming partner. Consequently, audit partner rotation is likely to yield effects that are second-order relative to effects of audit firm rotation. As researchers try to identify increasingly subtle economic effects, increasing the power of the empirical tests becomes more important. Chi et al. (2009) add to the emerging literature on the effects of audit partner rotation. This literature is more limited than the audit firm rotation literature because of limited availability of data on the identity of lead audit partners running specific engagements. Australia and Taiwan are exceptions where the lead part- ner(s) can be identified. On the basis of an analysis of a sample of clients in 1995 before audit partner rotation became mandatory in Australia, Carey and Simnett (2006) conclude that audit quality suffers when the audit partner’s tenure exceeds seven years. Specifically, when the partner’s tenure exceeds seven years, finan- cially distressed clients are less likely to receive a going-concern qualification, and there is also some evidence that the client is less likely to report earnings that just miss breakeven. (They also conclude that this deterioration in audit quality is attributable to non–Big 6 auditors.) However, Carey and Simnett (2006) find no evidence that signed or absolute abnormal accruals are associated with audit partner tenure. In contrast, Chen et al. (2008) conclude that audit quality improves with audit partner tenure. In the 1990 – 2001 period, before audit partner rotation became mandatory in Taiwan, Chen et al. (2008) conclude that both absolute performance-matched abnormal accruals and income-increasing performance- matched abnormal accruals decrease with audit partner tenure. 4 396 Contemporary Accounting Research CAR Vol. 26 No. 2 (Summer 2009) Chi et al. (2009) contribute by examining the effect of mandatory audit partner rotation in Taiwan. They conclude that mandatory partner rotation does not improve audit quality and may, in fact, impair audit quality relative to the same client firm in the year before mandatory rotation. Specifically, they find that abso- lute and income-increasing performance-matched abnormal accruals are larger for clients experiencing mandatory partner rotation than for the same clients in the year before the mandatory rotation. It is comforting that this inference is similar to the evidence in Chen et al. 2008, which is also based on data from Taiwan, albeit during an earlier period, when partner rotation was voluntary rather than manda- tory. However, Chi et al. find no difference between mandatory rotation clients’ abnormal accruals and those of (a) other clients not subject to mandatory rotation or (b) clients whose audit partners rotated voluntarily (before rotation was required). Also, they find that the ERC is higher for mandatory rotation clients than the voluntary rotation benchmark group, but there is no difference between the mandatory rotation group and the other two benchmark groups. 3. Research design and statistical power Chi et al. (2009) is almost a “no-results” paper. This does not mean the paper should not be published. Greenwald (1975) explains that the well-documented edi- torial bias against publishing papers that fail to reject the null hypothesis delays the acquisition of knowledge by fostering the publication of studies whose results are true, but of limited generalizability. We would add that unwillingness to publish “no-results” papers impedes science in two additional ways: (a) by obfuscating research initiatives that yield unexpected results, thereby failing to prevent other researchers from going down the same less-than-fruitful path; and (b) by giving researchers dysfunctional incentives to torture the data until those data confess something (anything!) with a p- value of 0.05 or better. Lindsay (1994) provides evidence that editorial “bias against the null” occurs in the accounting discipline. Bamber, Christensen, and Gaver (2000, 124) further argue that in combination with the bias against publishing replications (which is more extreme in the field of accounting than in hard sciences, where replication is the norm), editorial bias against the null “can lead to a situation where the first published studies are more likely to reject the null, and these initial studies have a disproportionate effect on subsequent research due to the bias against publishing replications”. For these rea- sons, we argue that it is vitally important to publish “no-results” studies. That said, it is equally important for researchers to make the case that the study’s empirical tests are powerful enough to detect an economically material effect, should one exist. Above, we have explained why any effect of audit partner rotation is likely to be modest. Detecting an effect of modest size requires powerful tests. Two inescapable features of the Chi et al. 2009 setting likely reduce the power of the study’s tests. First, focusing on the effects of mandatory audit partner rotation (which is, after all, the auditor rotation question of greatest practical import) necessarily limits variation in the audit partner tenure variable. Chen et al. (2008) consider tenure of 5 years or less to be short, and tenure exceeding 10 years to be long. Carey and Discussion of “Mandatory Audit Partner Rotation and Audit Quality” 397 CAR Vol. 26 No. 2 (Summer 2009) Simnett (2006) conclude that audit partner tenure must exceed 7 years to adversely affect audit quality. But after mandatory rotation, Taiwanese auditors cannot serve more than 5 consecutive years. As a result, while average audit partner tenure is shortest in the mandatory partner rotation sample (1.5 years), average partner ten- ure is not much longer in two of the three benchmark nonrotation samples (2.3 years for other clients whose partners are not required to rotate and 2.7 years for clients whose partners voluntarily rotated). 5 Should we expect an audit partner ten- ure difference of 0.8 to 1.2 years to materially affect audit quality? With the benefit of hindsight, it is perhaps not surprising that there are no consistently significant differences between the mandatory rotation sample and these two benchmark non- rotation samples. In contrast, the primary significant difference that Chi et al. (2009) find is between the mandatory rotation sample and the benchmark sample where partners have the longest tenure (4.8 years) — that is, the mandatory rota- tion sample in the prior year. The second feature that likely reduces the power of the study’s tests is that the archival audit literature sorely lacks sharp proxies for audit quality. As Chi et al. (2009) carefully note, proxies such as abnormal accruals and earnings response coefficients may be among the more popular measures used to date, but they are nonetheless noisy measures of earnings quality, much less audit quality. The next section expands on this issue. 4. Wanted: Sharper proxies for audit quality Audit quality is unobservable. Most of the literature draws inferences about audit quality based on traditional, noisy measures of earnings quality. Although the two are related, they are by no means isomorphic. A high-integrity client can produce high-quality earnings whether or not the audit is of high quality. Conversely, even a high-quality audit cannot be expected to identify or adjust all of a client’s low-quality reporting choices, so a high-quality audit does not guarantee high-quality earnings. 6 Two of the most popular measures of “audit quality” are abnormal accruals (e.g., Johnson et al. 2002; Myers et al. 2003; Carey and Simnett 2006; Chen et al. 2008), and ERCs (e.g., Ghosh and Moon 2005). However, the popularity of such measures does not mean that they are good proxies for earnings quality, much less audit quality. To their credit, Chi et al. (2009) forthrightly acknowledge some of the limitations of these proxies. In hopes of motivating future researchers to develop sharper and more refined proxies for audit quality, we expand on their discussion. Abnormal accruals are residuals from models of nondiscretionary (or normal) accruals. The explanatory power of these models rarely exceeds 30–35 percent. Is it plausible that most of the variation in aggregate accruals is due to inappropriate earnings management that auditors could have detected, but failed to eliminate? That is the implication of using abnormal accruals to evaluate audit quality. Given the modest explanatory power of these models and other well-known concerns reviewed in McNichols 2000, it is not surprising that Jones, Krishnan, and Melendrez (2008) find that most measures of abnormal accruals (including performance- matched abnormal accruals such as those in Chi et al. 2009) do not have any incremental explanatory power (beyond total accruals) in explaining extremely 398 Contemporary Accounting Research CAR Vol. 26 No. 2 (Summer 2009) low-quality earnings as evidenced by Securities and Exchange Commission (SEC) charges of fraudulent reporting. That is, the extraction of the “abnormal” portion of accruals does not create a sharper indicator of alleged fraudulent reporting than the use of total accruals. With respect to ERCs, using share prices as a benchmark for evaluating earn- ings quality implies that share prices correctly measure the firm’s fundamental value, and thus earnings numbers that more closely reflect the information impounded in prices are of higher quality. However, it appears that prices can diverge from fun- damental value for extended periods (e.g., Lee 2001, 2008). If markets can be fooled by complex or fraudulent financial reporting, share prices are a questionable benchmark for identifying the quality of those financial reports — particularly low- quality reports. 7 Finally, we find it a bit hard to believe that rotating one audit part- ner would materially affect the extent to which a client’s reported earnings reflect the same information priced by the market. While an extensive stream of research has used abnormal accruals and ERCs as proxies for audit quality, in the next section we argue that the archival audit liter- ature has now progressed to the point where it is time to move on. 5. Conclusions and suggestions for future research Chi et al. (2009) identify a pressing question of real practical import: Does manda- tory audit partner rotation improve audit quality? Because audit partners are not identified in North American audit reports, the authors capitalize on available evidence from Taiwan. This is an excellent use of international data to address a ques- tion of real importance in North America, but one that we cannot answer in a North American context. The empirical work is consistent with current standard method- ology. Although we take issue with the paper’s proxies for audit quality, this concern applies to the voluminous stream of archival research on audit quality, of which Chi et al. 2009 is simply one example. Turning to directions for future research, both the audit firm rotation literature and the audit partner rotation literature focus on the net effects of rotation, rather than separately identifying the costs and benefits of rotation. Future research could contribute by providing more insight into the specific costs and benefits. Because rotation creates significant costs, it would be useful to know, for example, whether and to what extent audit partner rotation leads to any measurable benefits at all. With the benefit of hindsight, it seems apparent that the magnitude of any effects of mandatory audit partner rotation would be modest. Particularly given the limited cross-sectional variation in audit partner tenure and the noisy proxies for audit quality, it is not surprising that the study does not find much empirical evi- dence of an effect. Future researchers who find themselves with a project where the results fail to reject the null would do well to heed Greenwald’s 1975 advice on gracefully failing to reject the null hypothesis. For example, he discusses the importance of provid- ing convincing evidence that the empirical proxies have construct validity, and he also suggests providing evidence on the size of an effect that the empirical analysis would be able to detect. The idea is to demonstrate that the study fails to reject the Discussion of “Mandatory Audit Partner Rotation and Audit Quality” 399 CAR Vol. 26 No. 2 (Summer 2009) null, not because the tests are insufficiently powerful, but rather because any effect that exists is economically immaterial. Cready and Mynatt (1991) provide an excellent illustration in their study of the price and trading volume reactions to the release of firms’ annual reports (as distinct from earnings announcements). After failing to reject the null hypothesis that there is no price reaction to the release of corporate annual reports, Cready and Mynatt (1991) provide simulation evidence demonstrating that if such a price reaction does exist, it is too small to be eco- nomically consequential, because their methods would identify a reaction of any meaningful magnitude. Finally, our primary suggestion for future research is that as archival audit research matures, it becomes increasingly necessary to move beyond the “usual suspect” proxies borrowed from archival financial accounting research. As research- ers address increasingly refined questions, the magnitude of hypothesized effects is likely to become more subtle and the relations more complex, which requires more powerful empirical methods, particularly sharper proxies of key constructs. Development of sharper measures of audit quality would provide a major breakthrough. This will require imagination and creativity. While this is a tall order, researchers have cleverly developed and validated empirical proxies for other unobservable theoretical constructs that were once thought to be unmeasurable. For example, Barron, Kim, Lim, and Stevens (1998) use properties of analysts’ forecasts to develop proxies for uncertainty, consensus, common information, and private information. Matsumoto (2002) develops a measure of managers’ total guidance of analysts, which Wang (2007) refines to develop a measure of managers’ private guidance (before Regulation Fair Disclosure). Development of more refined proxies for audit quality would in our opinion provide an equally, if not more, significant contribution. Sharper proxies for audit quality will undoubtedly capitalize on the institu- tional features of the audit environment. For example, rather than focusing on aggregate accruals, researchers could focus on specific accruals that (a) are eco- nomically significant, (b) are not well explained in accompanying disclosures, (c) are susceptible to manipulation, and (d) whose manipulation auditors are likely to be able to detect. Examples include loan loss provisions of financial institutions and insurers’ loss reserves. Field studies such as the Nelson, Elliott, and Tarpley 2002 analysis of managers’ decisions about how to attempt earnings management and auditors’ decisions whether to respond by requiring adjustments to the finan- cial statements may be helpful in identifying specific powerful contexts. DeAngelo (1981, 186) defines the quality of audit services as the “market assessed joint probability that a given auditor will both (a) discover a breach in the client’s accounting system, and (b) report the breach”. Researchers might incorporate factors related to auditors’ ability to detect questionable reporting practices, such as auditor industry specialization, auditor or audit firm experience, industry-specific characteristics, client complexity, or audit firm alumni in client firm top manage- ment. Factors related to auditors’ willingness to report or correct a questionable reporting practice, such as the importance of the client, the effect of the practice on 400 Contemporary Accounting Research CAR Vol. 26 No. 2 (Summer 2009) reaching an earnings target, the quality of the client’s corporate governance, or other measures of independence, should also affect audit quality. Researchers may be able to develop better measures of audit quality by combin- ing several indicators into an index score. In addition to the indicators mentioned above, research might also incorporate the client’s propensity to rarely just miss and often just beat key earnings benchmarks, client financial restatements, allega- tions of client fraudulent reporting, clients filing for bankruptcy after receiving a clean opinion, the auditor’s propensity to issue qualified opinions, and lawsuits against auditors. The chosen characteristics could be combined into a single index score, and validated in a manner analogous to Botosan’s 1997 approach to develop- ing and validating a disclosure index. Finally, researchers may be able to identify direct measures of audit quality in certain contexts, such as nonprofit firms’ compliance with GAAP disclosure requirements (Krishnan and Schauer 2000), errors in insurance firms’ estimates of claim loss reserves (Petroni and Beasley 1996), or ratings developed by government agencies (Deis and Giroux 1992; Copley, Doucet, and Gaver 1994). Endnotes 1. For example, the Conceptual Framework for AICPA Independence Standards recognizes the rotation of senior audit team members as an independence safeguard (American Institute of Certified Public Accountants [AICPA] 2006). 2. Reflecting differences between audit firm tenure and auditor tenure, Bamber and Iyer (2007) also find that individual auditors’ tenure is associated with greater identification with the client, which in turn increases individual auditors’ acquiescence to the client’s preferences. 3. We raise the issue of superficial audit partner rotation in Taiwan because this is a major difference between the Taiwanese environment and the U.S. environment, where the audit partner must remain rotated off for five years. In their empirical analysis, Chi et al. (2009) appropriately exclude clients where one audit partner rotated off in 2003 and rotated back on in 2004. The proportion of clients with this type of superficial rotation is about 10 percent of their preliminary sample, which is, however, much lower than the proportion Chen et al. (2008) report for 2003 and 2004. It appears that the voluntary rotation benchmark group could include superficial rotation, if audit firms engaged in superficial partner rotation in anticipation of the requirements for mandatory rotation. 4. Chen et al. (2008, 419) recognize that the differences between their inferences and Carey and Simnett’s 2006 inferences could be due to differences in research design (e.g., 1-year versus 12-year sample periods) or to institutional differences between Australia and Taiwan. Chen et al. also find that longer audit firm tenure is associated with higher earnings quality, similar to results found in studies using U.S. data. Carey and Simnett (2006) do not investigate the relation between audit firm tenure and earnings quality. However, Chen et al. (2008) argue that because audit partner tenure is strongly associated with audit firm tenure, Carey and Simnett’s (2006) evidence that longer audit partner tenure is associated with lower quality earnings suggests that Discussion of “Mandatory Audit Partner Rotation and Audit Quality” 401 CAR Vol. 26 No. 2 (Summer 2009) longer audit firm tenure may also be associated with lower quality earnings in Australia, contrary to results based on U.S. data. 5. Part of the reason for the modest difference in audit partner tenure arises because each engagement has two audit partners. Because the data do not distinguish the lead partner, the authors take the average of the two partners’ tenure. As a result, partner tenure can actually be shorter in the benchmark samples where rotation is not required or where there was voluntary rotation. For example, client A has two partners who have each served two years, so the average partner tenure is two years, and the partners are not required to rotate. In contrast, client B is required to rotate and has one new partner and one partner who has served four years, so the average partner tenure is two and a half years. 6. For example, even a high-quality auditor cannot require adjustments for suspected earnings management that the client achieved by structuring transactions to meet specific and precise accounting standards. 7. Similarly, Kothari (2001, 132) notes that “the reasons why maximizing the earnings’ correlation with stock returns is desirable are not well articulated or proven logically”. References American Institute of Certified Public Accountants (AICPA). 2006. ET Section 100.01 Conceptual Framework for AICPA Independence Standards. New York: AICPA. Bamber, L., T. Christensen, and K. Gaver. 2000. Do we really “know” what we think we know? A case study of seminal research and its subsequent overgeneralization. Accounting, Organizations and Society 25 (2): 103–29. Bamber, E., and V. Iyer. 2007. Auditors’ identification with their clients and its effect on auditors’ objectivity. Auditing: A Journal of Practice & Theory 26 (2): 1–24. Barron, O., O. Kim, S. Lim, and D. Stevens. 1998. Using analysts’ forecasts to measure properties of analysts’ information environment. The Accounting Review 73 (4): 421–33. Botosan, C. 1997. Disclosure level and the cost of equity capital. The Accounting Review 72 (3): 323–49. Carey, P., and R. Simnett. 2006. Audit partner tenure and audit quality. The Accounting Review 81 (3): 653–76. Chen, C., C. Lin, and Y. Lin. 2008. Audit partner tenure, audit firm tenure, and discretionary accruals: Does long auditor tenure impair earnings quality? Contemporary Accounting Research 25 (2): 415–45. Chi, W., H. Huang, Y. Liao, and H. Xie. 2009. Mandatory audit partner rotation, audit quality, and market perception: Evidence from Taiwan. Contemporary Accounting Research 26 (2): 359–91. Copley, P., M. Doucet, and K. Gaver. 1994. A simultaneous equations analysis of quality control review outcomes and engagement fees for audits of recipients of federal financial assistance. The Accounting Review 69 (1): 244–56. Cready, W., and P. Mynatt. 1991. The information content of annual reports: A price and trading response analysis. The Accounting Review 66 (2): 291–312. DeAngelo, L. 1981. Auditor size and audit quality. Journal of Accounting and Economics 3 (3): 183–99. 402 Contemporary Accounting Research CAR Vol. 26 No. 2 (Summer 2009) DeFond, M., and J. Francis. 2005. Audit research after Sarbanes-Oxley. Auditing: A Journal of Practice & Theory 24 (Supplement): 5–30. Deis, D., and G. Giroux. 1992. Determinants of audit quality in the public sector. The Accounting Review 67 (3): 462–79. Ghosh, A., and D. Moon. 2005. Auditor tenure and perceptions of audit quality. The Accounting Review 80 (2): 585–612. Greenwald, A. 1975. Consequences of prejudice against the null hypothesis. Psychological Bulletin 82 (1): 1–20. Johnson, V., I. Khurana, and J. Reynolds. 2002. Audit-firm tenure and the quality of financial reports. Contemporary Accounting Research 19 (4): 637–60. Jones, K., G. Krishnan, and K. Melendrez. 2008. Do models of discretionary accruals detect actual cases of fraudulent and restated earnings? An empirical analysis. Contemporary Accounting Research 25 (2): 499–531. Kothari, S. 2001. Capital markets research in accounting. Journal of Accounting and Economics 31 (1–3): 105–231. Krishnan, J., and P. Schauer. 2000. The differentiation of quality among auditors: Evidence from the not-for-profit sector. Auditing: A Journal of Practice & Theory 19 (2): 9–25. Lee, C. 2001. Market efficiency and accounting research: A discussion of “Capital market research in accounting” by S. P. Kothari. Journal of Accounting and Economics 31 (1–3): 233–53. Lee, C. 2008. The magic of markets: Market efficiency and capital market research in accounting. http://fars.org/CharlesLee2008AAA.ppt. Lindsay, R. 1994. Publication system biases associated with the statistical testing paradigm. Contemporary Accounting Research 11 (1): 33–57. Mansi, S., W. Maxwell, and D. Miller. 2004. Does auditor quality and tenure matter to investors? Evidence from the bond market. Journal of Accounting Research 42 (4): 755–93. Matsumoto, D. 2002. Management’s incentives to avoid negative earnings surprises. The Accounting Review 77 (3): 483–514. McNichols, M. 2000. Research design issues in earnings management studies. Journal of Accounting and Public Policy 19 (4–5): 313–45. Myers, J., L. Myers, and T. Omer. 2003. Exploring the term of the auditor-client relationship and the quality of earnings: A case for mandatory auditor rotation? The Accounting Review 78 (3): 779–99. Nelson, M., J. Elliott, and R. Tarpley. 2002. Evidence from auditors about managers’ and auditors’ earnings management decisions. The Accounting Review 77 (Supplement): 175–202. Petroni, K., and M. Beasley. 1996. Errors in accounting estimates and their relation to audit firm type. Journal of Accounting Research 34 (1): 151–71. Sarbanes-Oxley Act of 2002. 2002. Pub. Law no. 107-204, 116 Stat. 745. Wang, I. 2007. Private earnings guidance and its implications for disclosure regulation. The Accounting Review 82 (5): 1299–332. . (Summer 2009) pp. 393–402 © CAAA doi:10.1506/car.26.2.3 Discussion of Mandatory Audit Partner Rotation, Audit Quality, and Market Perception: Evidence from Taiwan * E. MICHAEL BAMBER, . arguments that the costs of audit firm rotation (in Discussion of Mandatory Audit Partner Rotation and Audit Quality” 395 CAR Vol. 26 No. 2 (Summer 2009) terms of lost client-specific knowledge). rotating audit firms versus audit partners Auditor rotation can be accomplished by changing either the audit firm or the audit partner. Most of the evidence on the effects of auditor rotation on audit