Changes in audit quality and audit fees as a consequence of
audit partner identification and differences among
litigious and non-litigious industries
Jim Bijvoet 10194215 20-06-2015
Word count: 15,919
MSc Accountancy & Control, variant Accountancy Amsterdam Business School
Faculty of Economics and Business, University of Amsterdam Supervisor: dr. A. Sikalidis
Statement of Originality
This document is written by student Jim Bijvoet who declares to take full responsibility for the contents of this document. I declare that the text and the work presented in this document is original and that no sources other than those mentioned in the text and its references have been used in creating it. The Faculty of Economics and Business is responsible solely for the supervision of completion of the work, not for the contents.
Abstract
This thesis investigates are what the consequences of requiring the audit partner to sign the audit report, which ensures that the auditor is indentified by name. The Public Company Accounting Oversight Board (PCAOB) is currently considering to mandate this requirement in the United States. In 2009, it was already introduced in the United Kingdom. Therefore, a sample of United Kingdom companies is used to compare the years before and after the implementation of the regulation to find out what is the impact of the audit partner identification. A database research is performed to find out if there are changes in audit quality and audit fees and if these changes differ among litigious and non-litigious industries. The audit quality, as measured by four components, does not seem to increase. In contrast, the amount of audit fees does increase after the auditor is required to sign the audit report. As a result, it does not seem desirable to introduce this requirement in the United States since the intended effect of increased audit quality is not achieved, while the costs of the audit do increase. The effect in terms of audit quality does not vary among litigious and non-litigious industries. Audit fees, on the other hand, increase more for companies operating within non-litigious industries as compared to firms in litigious industries.
Table of contents
1. Introduction 5
1.1 Background information 5
1.2 Research Question 7
1.3 Motivation of the study 8
1.3.1 Motivation from an academic point of view 8 1.3.2 Motivation from a societal point of view 10
2. Literature Review and Hypotheses 11
2.1 Changes in audit quality as a consequence of audit partner identification 11 2.2 Changes in audit fees as a consequence of audit partner identification 13 2.3 Differences in audit quality between litigious and non-litigious industries 14 2.4 Differences in audit fees between litigious and non-litigious industries 16
2.5 Hypotheses 18
3. Methodology 19
3.1 Research method and sample selection 19
3.2 Empirical models 20
3.2.1 Abnormal accruals models 21
3.2.2 Small earnings increase models 21
3.2.3 Earnings informativeness models 22
3.2.4 Audit opinion models 23
3.2.5 Audit fee models 24
4. Empirical Results 24
4.1 Abnormal accruals results 26
4.2 Small earnings increase results 28
4.3 Earnings informativeness results 29
4.4 Audit opinion results 30
4.5 Audit fee results 32
5. Discussion and Conclusion 33
5.1 Audit quality change as a result of the audit partner signature requirement 34 5.2 Audit fees change as a result of the audit partner signature requirement 35 5.3 Audit quality change differences between litigious and non-litigious industries 36 5.4 Audit fees change differences between litigious and non-litigious industries 37
5.5 Conclusion 37
References 38
1. Introduction
1.1 Background information
Firms publish their financial statements on an annual basis. These statements are checked for completeness and correctness by an auditing firm. One audit partner, along with his audit team, performs the financial statement audit and gives his verdict on the appropriateness of the statement. Usually, the audit partner is not identified by name. Therefore, the readers of the financial statements cannot obtain any additional information about this person and this makes sure that there is no impact on the perceived quality of the audit. However, over the past decade, publishing the name of the audit partner and therefore identifying this person has become a requirement in several countries, including the United Kingdom in 2009 (Carcello & Li, 2013, p. 1512). Audit partners have to sign the audit report in order to inform the readers of the financial statements about their identity.
This requirement has various consequences, mainly in the form of changed audit quality and audit fees. Audit quality might change because firms might want to improve the quality of the audit partners they choose to check their financial statements. When the audit partner can be identified, investors and other stakeholders would prefer the financial statement audits to be performed by a competent person. Therefore, companies would want to hire better audit partners, which would improve the audit quality (King, Davis & Mintchik, 2012, p. 534). In addition, when audit partners are identified by name, they are more accountable for their work. In particular, they would have more motivation to avoid negative consequences of their actions and are encouraged to perform well. This would result in higher audit quality because they would, for instance, perform additional audit procedures, which would result in increased performance (DeZoort, Harrison & Taylor, 2006, p. 376). Audit fees could also be different when audit partner’s identity is known. If companies would decide to choose another engagement partner, this person might charge a higher amount for his services. Therefore, this might result in higher audit fees which have to be paid. Also, if audit partners would perform additional audit procedures and spend more hours to audit a company’s financial statements, this extra work would also increase the audit fees which have to be paid (Carcello & Li, 2013, p. 1518).
The Public Company Accounting Oversight Board (PCAOB) is currently considering mandating the disclosure of the engagement partner’s identity in the United States (Yen, Chang & Chen, 2013, p. 230). The main reason for implementing the mandatory signature of the audit partner would be to increase the quality of the financial statement audits performed by this person. Some research has been conducted over the past years in order to determine the relationship between requiring audit partner identification and subsequent changes in
both audit quality and audit fees. The studies which were published mostly showed mixed results in terms of changes in audit quality. Audit fees, on the other hand, seemed to increase in most cases.
The study conducted by Carcello and Li (2013) researched the effects of the required audit partner signature on audit quality and audit fees in the United Kingdom since its implementation in 2009. In order to investigate these effects, a sample was used consisting of companies from the United Kingdom in the years prior to and after the implementation of the audit partner identification. Audit quality was measured by taking into account four variables that represented audit quality and audit fees were measured by comparing the fees paid before and after the requirement. It was found that both audit quality and audit fees increased because of the mandatory audit partner identification (Carcello & Li, 2013, p. 1511). The audit partner signature requirement benefited shareholders and other stakeholders in terms of audit quality, but higher fees were paid in order to achieve this result. According to the researchers, the results of their study are relevant for the decision to be made by the PCAOB about implementing the requirement in the United States. They state that the United Kingdom is in many ways similar to the United States and that because of this, the same results would apply in both countries.
Another article which discusses the change in audit quality as a consequence of audit partner identification is based on research performed in The Netherlands. Blay, Notbohm, Schelleman and Valencia (2014) investigated the effects of the mandatory audit partner signature implemented in the Netherlands in 2006. They used a sample of Dutch firms to find out if the audit quality had changed after the audit partner identification was required. This sample was compared to a sample consisting of United Kingdom companies before the implementation of the requirement in the United Kingdom in 2009, because of the similarities between both countries. Audit quality was measured by using several measures of earnings quality. It was concluded that the audit quality remained the same as before the implementation of the requirement (Blay et al., 2014, p. 172). Therefore, the authors stated that the mandatory signature of the audit partner did not have its intended effect in The Netherlands.
King et al. (2012) also investigated whether or not the mandatory audit partner signature increases audit quality. To research this matter, they did not use a sample of firm data, but rather focused on factors that potentially affect audit quality in appearance and audit quality in fact. In the paper, insights from three distinct academic frameworks are applied. Source credibility, accountability and the theory of affordances are taken into account to find out what the effect of the audit partner identification is on audit quality. The study shows that there is not enough evidence to suggest that a mandatory audit partner signature will lead to higher audit quality in fact. Audit quality in appearance, however, is
likely to increase after the implementation of the identification requirement (King et al., 2012, p. 554). The authors state that an increase in audit quality in appearance is only desirable when actual audit quality is higher, because otherwise the expectations of financial statement users about the audit quality would become too high.
Research performed by Yen et al. (2013) looked at the audit quality change as an effect of the mandatory audit partner signature in Taiwan. The identification of the audit partner has been a requirement in Taiwan since 1983. The paper employed a questionnaire survey in order to investigate the opinions of the readers of the audit report. Most of the people that took part in the survey agreed that the audit partner signature requirement would increase the accountability of auditors. They also thought that knowledge of the identity of the audit partner is important to financial statement users. However, less than half of the respondents agreed that an audit partner signature requirement would result in higher audit quality (Yen et al., 2013, p. 232). Because of this result, the authors concluded that a mandatory engagement partner signature would not have its intended effect on the audit quality.
1.2 Research Question
As mentioned before, Carcello and Li (2013) found that both audit quality and audit fees increased as a result of the requirement of the engagement partner to sign the audit report in the United Kingdom. This result provides clear and interesting evidence about the consequences of the mandatory identification of the audit partner. Following the outcome of this study, this thesis will specify these effects further by looking at company characteristics in more detail. This will be done by looking at the possible changes of the audit quality and audit fees among the United Kingdom’s litigious and non-litigious industries as a consequence of the audit partner’s signature requirement. Because of this, the research question of this thesis is:
“Have the audit quality and audit fees in the United Kingdom increased as a consequence of the audit partner’s signature requirement and do these results vary between litigious and non-litigious industries?”
By answering this research question, this thesis will not only provide additional evidence about the changes in audit quality and audit fees as a result of the audit partner signature requirement, but it will also show if these effects differ based on whether a company is in a litigious industry or not. In order to answer the question, data about United Kingdom companies will be used to find the effects of the audit partner identification on both types of
industries. The sample that will be used is similar to the one used in the study by Carcello and Li (2013). Also, the measurements of the theoretical constructs are mainly based on the measurements used in this study.
Because of the increased audit quality and audit fees found by Carcello and Li (2013), it is expected that similar results will be found in this study with regard to the audit partner identification. However, no research has been performed about the difference between litigious and non-litigious industries in terms of changes of audit quality and audit fees as a consequence of audit partner identification. The expectation is that the mandatory audit partner signature will result in a bigger audit quality increase in litigious industries than in non-litigious industries. The reason for this is that litigious industries usually have lower audit quality (Palmrose, 1988, p. 55) and this is therefore more likely to increase. Audit fees, on the other hand, are lower in non-litigious industries (Simunic & Stein, 1995, p. 119) and therefore a bigger increase is assumed to take place in these industries than in litigious industries.
1.3 Motivation of the study
Answering the research question of this thesis is useful for various reasons. There are several consequences of the mandatory signature of the audit partner with respect to audit quality and audit fees. Also, the possible differences between litigious and non-litigious industries might have certain effects. The results provided by this thesis can contribute to the scientific knowledge and multiple stakeholders might be interested in these results. There are several reasons, from both an academic and a societal point of view, why this is the case.
1.3.1 Motivation from an academic point of view
By investigating the effects of mandatory identification of the engagement partner on audit quality and audit fees, this thesis will make various contributions to the existing academic literature. In addition, by distinguishing between litigious and non-litigious industries, previously unknown consequences will be discovered. Therefore, from a scientific viewpoint, there are multiple reasons why the results would be interesting.
Firstly, there has not been a lot of research about the relationship between a mandatory audit partner signature and the consequences of this requirement on audit quality and audit fees. Because of this, there is only limited knowledge about this topic and limited evidence to support the findings of the research. Therefore, a contribution to the existing knowledge can be made by answering the research question of this thesis. In addition, the studies that investigated the relationship between audit partner identification and its
consequences only focused on the effects on all companies within a specific country or region. No distinction between various company characteristics has been made. The research of this thesis will focus on the differences between companies that operate in litigious and non-litigious industries, in order to find out whether the results might vary among those different industries. This has not been investigated before and therefore a research contribution can be made by extending the knowledge and understanding of the consequences of the audit partner signature requirement and the effect this has on litigious and non-litigious industries.
Another reason why this study contributes to the academic literature is that the measures of the audit quality are more reliable. As opposed to the studies that have been conducted before about the effects of audit partner identification on audit quality and audit fees, this thesis uses more models to measure audit quality. Most of the models which are used to measure the consequences of a mandatory audit partner signature on audit quality have been extracted from Carcello and Li (2013). In addition, models from several other studies have been used in order to assess audit quality. By using more models, the effects of the mandatory audit partner signature requirement on audit quality becomes clearer. The reason for this effect is that there is a greater chance of a correct prediction of the change in audit quality when multiple measures have been taken into account and if these all provide the same results. Therefore, this thesis makes a contribution by providing more reliable results about a change in audit quality as a consequence of the identification of the audit partner.
Furthermore, the study which is conducted is interesting from a scientific point of view because it makes a contribution to the knowledge about the change in audit fees as a result of mandatory audit partner identification. Little is known about this effect because not much research has focused on this topic. This far, only one study has examined the effects that a mandatory audit partner signature requirement has on audit fees. The study carried out by Carcello and Li (2013) investigated the effects on both audit quality and audit fees as a consequence of audit partner identification and it was shown that audit fees increased. Because this is the only study that examined this effect, this thesis can contribute to the existing academic literature by demonstrating the possible change in audit fees because of the identification of the audit partner. This will provide additional evidence to enhance the limited knowledge about this topic.
1.3.2 Motivation from a societal point of view
Mandating the audit partner to sign the audit report can also have consequences for society and its members. Several groups and individuals might be interested in what the effect of
audit partner identification is on audit quality and audit fees. Also, they would possibly want to know whether or not the results differ among litigious and non-litigious industries.
The PCAOB is currently considering to mandate the identification of the audit partner in the United States. This was also the main reason to research this topic for the study by Carcello and Li (2013, p. 1512). Making the engagement partner’s signature a requirement has some consequences for the audit quality and audit fees, which was shown in the mentioned study. Because of this, the results of this thesis might be useful for the PCAOB and their decision regarding whether or not to implement the audit partner identification requirement. This far, all the research that investigated this topic did not look at company characteristics like industry type, but looked at the effects of the signature requirement on audit quality and audit fees in general. This makes sure that the results found can possibly not be applied to all companies in a specific country. Therefore, this does not give a complete view of the effects that the requirement has on different types of companies. This thesis also looks at the effects of the audit partner identification based on whether or not the company operates in a litigious industry. It might be interesting for the PCAOB and standard setters from other countries to know what the consequences of a mandatory audit partner signature would be for the different types of industries, because this is currently not known.
The results of this thesis might also be interesting for audit firms and the audit partners they employ. The research which will be performed will show if the audit quality and audit fees will change as a result of the mandatory audit partner signature and if this varies among litigious and non-litigious industries. The audit firms and audit partners would want to know whether the requirement would result in better performances that they provide. If they would provide higher-quality financial statement audits, audit report users would have more confidence in their opinion, which would increase their reputation (Skinner & Srinivasan, 2012, p. 1759). In addition, audits of higher quality would prevent approving incorrect financial statements, which would reduce the chance of accounting scandals taking place (Kleinman, Lin & Palmon, 2014, p. 61). The change in audit fees would also be of interest to audit firms and their audit partners. If firms would be willing to pay them higher fees to enhance the audit quality of their financial statements, the income of the firms would rise and the salaries of the audit partners would also increase.
Another reason why answering the research question can be valuable is that companies can benefit from knowing and understanding the consequences of the implementation of the requirement, both in general and specifically for litigious and non-litigious industries. Firms would have a better understanding of whether or not the mandated audit partner identification would lead to increased audit quality of their financial statements and if they would have to pay higher audit fees in order to achieve this. Information about changes in audit quality and audit fees as a result of the requirement might be important for
firms to know. They would benefit from increased audit quality of their financial statements, mainly because financial statement users would have more confidence in the reliability of the financial information they provide (Skinner & Srinivasan, 2012, p. 1759). This increased audit quality might result in higher audit fees to be paid in order to attain this, which would increase the firm’s costs. Since there is only limited knowledge about the effects of the audit partner signature requirement, answering the research question can add additional evidence on this topic. In addition, because this thesis also examines the effects on litigious and non-litigious industries specifically, firms might obtain increased understanding of the changes in audit quality and audit fees for the type of industry they operate in.
Lastly, the research question and its answer would be of interest to the readers of the financial statements and the audit reports, which are the investors and other stakeholders of the company. The information provided by this thesis would benefit them mostly in terms of the possible increase of the audit quality, both in general and per industry type. Because the shareholders and all other stakeholders depend a lot on the financial information provided by the company, an increase in audit quality would benefit them (Francis & Schipper, 1999, p. 350). They make important decisions based on the information provided in the financial statements. If higher assurance about the correctness and completeness of this information can be given, there is an increased chance that better results follow from their decision-making. Also, the possible increase in audit fees might be an interesting consequence in their opinion. Because shareholders are the owners of the company, they would want to know whether the implementation of the audit partner signature requirement would result in higher annual costs for the firm because of increased audit fees.
2. Literature Review and Hypotheses
2.1 Changes in audit quality as a consequence of audit partner identification
One of the consequences of implementing the audit partner signature requirement is that audit quality can change as a result. Improving the audit quality is the main reason for the PCAOB to possibly implement the requirement in the United States (Blay et al., 2014, p. 173). They base their opinion on four principles. The first one is that audit partner identification would increase transparency in the audit process and thereby providing more information to readers of the audit report. Secondly, the signature requirement would increase the accountability of audit partners, which would mean they would perform better because they feel more accountable for their work. Another principle is that the identification would result in more responsibility of the audit partner for the audit quality. The final principle
is about the fact that firms would want to hire more competent audit partners, who would provide higher-quality financial statement audits. All of these consequences of the audit partner signature requirement would, according to the PCAOB, result in higher audit quality (King et al., 2012, p. 534). In order to find out whether these effects would take place in reality, the PCAOB asked for empirical studies to verify their assumptions (Yen et al., 2013, p. 232). As a consequence, research has been conducted over the past years to find out what the effect of the audit partner signature requirement is on audit quality. Multiple studies have been carried out, but the results that were found seem to vary.
The study conducted by Carcello and Li (2013) shows that the assumptions made by the PCAOB apply to the United Kingdom. Their study examined the effects of the mandatory audit partner signature, which was implemented in 2009, on audit quality. They looked at the years before and after the implementation to find out if a change had taken place. Since audit quality is unobservable, four different variables were used as a way to measure it: abnormal accruals, the likelihood of firms reporting a small earnings increase, earnings informativeness and the likelihood of auditors to issue a qualified audit opinion. The effects of the audit partner identification on these variables all showed that audit quality increased as a consequence of the signature requirement (Carcello & Li, 2013, p. 1542). The result of this study shows that there is evidence to confirm the expectations of the PCAOB. According to Carcello and Li (2013, p. 1543), the outcome of their research in the United Kingdom is relevant for the decision to be made by the PCAOB about implementing the requirement in the United States because of the similarities between both countries.
An opposing result was found by Blay et al. (2014) in a study that was carried out in The Netherlands. The authors investigated the change in audit quality as a consequence of the audit partner signature which was implemented in 2006. In order to measure audit quality, various measures of earnings quality were used. It was found that the mandatory audit partner signature did not have its intended effect in The Netherlands. This was the case because the audit quality did not change after the implementation of the requirement (Blay et al., 2014, p. 172). The authors state that the reason for this result could be that accountability levels and audit quality already were at high levels before the mandatory audit partner signature (Blay et al., 2014, p. 188). The result of this study is not in line with the assumptions of the PCAOB, which means that their expectations may not turn out to be true in reality.
Another study which investigated the effects of a mandatory audit partner signature on audit quality was carried out by King et al. (2012). As opposed to the other studies that were mentioned, the authors did not use a sample of firm data to research this topic. Instead, they focused on factors that potentially affect audit quality in appearance and audit quality in fact. The study applied insights from three distinct academic frameworks: source credibility,
accountability and the theory of affordances. The authors show that the impact of the audit partner identification on audit quality in fact remains unclear. An increase in audit quality in appearance, however, is likely to happen after a mandatory audit partner signature is implemented (King et al., 2012, p. 533). According to King et al. (2012, p. 554), an increase in audit quality in appearance is only desirable when actual audit quality is higher. The reason for this is that otherwise a gap between public perceptions and actual audit quality will appear or increase.
The final study that researched the possible change in audit quality as a consequence of audit partner identification was conducted by Yen et al. (2013). Their research was performed in Taiwan, where an audit partner signature has been mandatory since 1983. A questionnaire survey was employed to find out what the opinions were of the audit report users on the effects of the signature requirement. Most of the respondents agreed that mandatory audit partner identification would lead to increased accountability and responsibility of audit partners (Yen et al., 2013, p. 230). Less than half of the respondents, however, agreed that an audit partner signature requirement would result in higher audit quality. Because of this result, the authors concluded that identification of the audit partner does not have its intended effects on audit quality (Yen et al., 2013, p. 235).
In conclusion, a mandatory audit partner signature does not seem to have a clear impact on the audit quality. Some studies show an increase in audit quality, while others show it remained at the same level as before an identification requirement was implemented. Because of this, no clear conclusions can be drawn after analysing the results of the studies which were mentioned. There does not seem to be a universal impact and the effects of a mandatory signature are most likely to depend on various circumstances, such as the country in which the study takes place or the type of study that is conducted.
2.2 Changes in audit fees as a consequence of audit partner identification
A change in audit quality is not the only effect that a mandatory audit partner signature can have. According to the PCAOB, audit fees can also change when identification of the audit partner is implemented (Carcello & Li, 2013, p. 1518). A reason for this consequence could be that audit partners would work more hours to improve the quality of their work, because they feel more accountable for their actions. These additional hours would result in higher total audit fees which have to be paid (Bell, Landsman & Shackelford, 2001, p. 35). In addition, firms might prefer to employ a more qualified audit partner who has more experience (King et al., 2012, p. 534). This could increase the audit quality which is provided, but higher audit fees might have to be paid to hire this person. As was already mentioned, the PCAOB requested researchers to investigate whether or not their assumptions would
hold in real-world situations (Yen et al., 2013, p. 232). Despite this call for increased research, not much evidence has emerged to find out if there is a possible change in audit fees after a mandatory audit partner signature is implemented. Only one study was conducted to research this matter.
Carcello and Li (2013) tested the expectations of the PCAOB regarding changes in audit fees. They conducted research in the United Kingdom by looking at the consequences of the implementation of an audit partner signature requirement in 2009. Audit fee amounts of the years before and after the requirement was introduced were compared to each other to find out if a change had taken place. The authors showed that the audit partner signature requirement resulted in higher audit fees: the amount of audit fees that was paid by firms increased by 13.2 percent (Carcello & Li, 2013, p. 1532). One of the reasons they provide for this result is that increased accountability causes audit partners to work more hours, resulting in higher audit fees. They state, however, that this increased effort may be unnecessary and that it could result in audit inefficiency, because more time is needed to complete the audit tasks. Furthermore, it was shown that higher audit fees were charged by Big 6 auditors, implying that more competent audit partners demand higher audit fees (Carcello & Li, 2013, p. 1532). These results confirm the assumptions that the PCAOB had about possible changes in audit fees as a consequence of audit partner identification. Because of the similarities between the United Kingdom and the United States, the outcome of this study is relevant for the decision that the PCAOB has to make.
To conclude, the amount of audit fees charged seems to increase as a consequence of the implementation of a mandatory audit partner signature. However, there is not much evidence to support this effect. Only one study has been carried out to provide results about the changes in audit fees as a consequence of audit partner identification. Because of this, the amount of proof is rather weak and this makes sure that there is no certainty about the existence of the effect. In order to get more verification, additional studies will need to be conducted to find out whether or not this effect seems to be true in reality.
2.3 Differences in audit quality between litigious and non-litigious industries
The research question of this thesis does not only look at whether audit quality changes as a result of an audit partner signature requirement, but also looks at whether this possible change differs among litigious and non-litigious industries. Different consequences might occur because the characteristics of both industry types vary. As was shown in previous paragraphs, several studies have been conducted to find out the effect of a mandatory audit partner signature on audit quality. No research has, however, investigated whether or not the audit quality changes vary between litigious and non-litigious industries. Because of this,
studies that examined the difference in audit quality between both types of industries will be discussed instead. Although the effect of an audit partner identification requirement is not taken into account, these studies provide insight into the differences between the industry types in terms of the audit quality that is provided.
Research carried out by Palmrose (1988) investigated whether auditors with relatively low litigation activity provide higher quality audits. In order to find out if this assumed effect is correct, a sample of legal cases involving independent auditors was used. These cases took place between 1960 and 1985 in the United States. The study found that non-Big Eight firms had higher litigation activity than Big Eight firms (Palmrose, 1988, p. 55). According to the author, this result shows that higher audit quality is found among audit companies that face less litigation concerns. This is because Big Eight firms are generally perceived to provide higher audit quality compared to other audit firms (Palmrose, 1988, p. 72). The outcome of this study shows that the quality of the financial statement audit is lower in industries with higher litigation risk, which means that there is lower audit quality in litigious industries.
Sun and Liu (2011) conducted a similar study to find out if audit quality differs among litigious and non-litigious industries. They investigated this matter by looking at whether or not litigation risk is related to the higher audit quality of big auditors. In order to research this topic, they collected financial statement data from United States listed companies from the period 1988-2006. As a way to measure audit quality, earnings management was used. The researchers found that the difference in audit quality between Big N and non-Big N auditors is more evident when litigation risk is high (Sun & Liu, 2011, p. 313). The higher audit quality that is provided by Big N firms therefore depends on the level of litigation concerns. This suggests that audit quality is lower in litigious industries and that audit quality in non-litigious industries is higher.
A study carried out by Venkataraman, Weber and Willenborg (2008) investigated whether audit quality is higher in a litigious or non-litigious industry. To research this topic, the authors used a setting that involved companies that had their Initial Public Offering. The sample consisted of firms that went public between the beginning of 2000 and the end of 2002. In order to measure audit quality, five different measures of abnormal accruals were used. The result of the study showed that audit quality was higher in a higher-litigation regime (Venkataraman et al., 2008, p. 1315). According to the researchers, this is consistent with the effect that an increase in litigation exposure should have on the incentives of the auditor. As opposed to the studies which were mentioned earlier, this study shows that audit quality in litigious industries is higher compared to non-litigious industries.
Research performed by Heninger (2001) examined the relation between earnings management and litigation. The study investigates the effect that reporting more income-increasing abnormal accruals has on litigation risk. Some researchers use abnormal accruals
as a proxy of audit quality. Therefore, the results of this study show what the relation is between audit quality and litigation. In order to research this, the author uses a sample of lawsuits brought by shareholders against auditors. Heninger (2001, p. 111) shows that litigation is positively associated with higher abnormal accruals, which represents low audit quality. This finding is consistent with stakeholders holding auditors responsible for failing to prevent the publication of incorrect financial information (Heninger, 2001, p. 124). The result of this study shows that litigation is related with low audit quality. This means that audit quality is lower in litigious industries.
In conclusion, audit quality seems to be lower in litigious industries. Although not all research supports this result, this is the main finding of most studies that have been conducted about this topic. The main interest of this thesis regarding this issue is if the changes in audit quality caused by audit partner identification differ among litigious and non-litigious industries. However, no research has been carried out about this possible effect. Because of this, the difference between litigious and non-litigious industries in terms of audit quality was examined and clearly showed that this is higher in non-litigious industries.
2.4 Differences in audit fees between litigious and non-litigious industries
In addition to the topics that were examined in the previous paragraphs, this thesis also looks at the differences between litigious and non-litigious industries in terms of audit fee changes as a result of audit partner identification. Because of the different characteristics of both types of industries, they are likely to provide different effects. Studies about the effect of an audit partner signature requirement on audit fees were examined earlier. However, no research considered whether or not changes in audit fees were different among litigious and non-litigious industries. Therefore, this paragraph will discuss studies that investigated the audit fee differences between both industry types. While the consequences of a mandatory audit partner signature are not taken into consideration, these studies provide insight into how audit fees differ among both industries.
Simunic and Stein (1995) perform research that aims at investigating the relationship between litigation risk and audit pricing. They examined whether audit fees were higher when there was increased litigation risk. In order to find useful results for their study, the authors used a sample of audits performed by Big 6 auditors in the United States. The evidence of the research suggests that audit fees are different when litigation risk varies: higher fees are paid when there are increased litigation concerns (Simunic & Stein, 1995, p. 119). According to the researchers, this shows that auditors are compensated for the increased lawsuit risks they face by higher fees that cover the costs of litigation. The outcome of this study clearly shows that higher audit fees are paid in litigious industries.
A study carried out by Seetharaman, Gul and Lynn (2002) examined the same topic and found similar results. They attempted to find out if a relationship between litigation risk and audit fees existed. In order to research this, the study looked at companies from the United Kingdom that were trading on markets in the United States. This combination was chosen because the United States markets represented a litigious environment, while information about audit fees was publicly disclosed for United Kingdom companies (Seetharaman et al., 2002, p. 113). The study found that audit fees reflected risk differences across liability regimes. Auditors charge significantly higher audit fees when their clients are operating in a litigious environment (Seetharaman et al., 2002, p. 113). This result shows that audit fees are higher in litigious industries than in non-litigious industries.
Research performed by Pratt and Stice (1994) also investigated whether audit fees are different when there is increased litigation risk. To test this, the authors asked managers and partners from Big 6 accounting firms to respond to a fictitious case that involved a potential audit client. The results showed that litigation risk increased audit fees (Pratt & Stice, 1994, p. 655). This effect happened for two reasons. Higher litigation risk caused additional auditing to be performed, which in turn led to higher audit fees. In addition, a premium was charged to cover possible future litigation losses (Pratt & Stice, 1994, p. 640). The study shows that increased litigation risks cause higher audit fees to be paid. Therefore, the outcome of this study is similar to research mentioned earlier: it suggests that higher audit fees are charged in litigious industries.
Choi, Kim, Liu & Simunic (2009) conducted a study to find out if auditors charge higher audit fees for firms that are cross-listed in countries with strong legal regimes. By examining this issue, the study shows whether a litigious environment results in higher audit fees to be demanded. In order to research this, the study uses a large sample of non-United States firms that are cross-listed in other countries. Choi et al. (2009, p. 1429) show that comparatively higher audit fees are charged by auditors for firms that are operating in a strong legal regime. The reason for this is that shareholders are willing to pay additional audit fee premiums to increase the audit quality that is provided (Choi et al., 2009, p. 1461). Similar to the studies which were mentioned before, this study shows that audit fees are higher in litigious environments.
To conclude, it seems clear that audit fees are higher in litigious industries than in non-litigious industries. All the studies that were discussed showed that this was the case. This thesis wants to investigate if changes in audit fees as a result of an audit partner signature requirement vary between litigious and non-litigious industries. There is, however, no research that has studied this possible difference. Therefore, the audit fee differences between litigious and non-litigious industries were examined in this paragraph and it was clearly shown that audit fees are higher in litigious industries.
2.5 Hypotheses
The research question of this thesis consists of several different components, which all have to be examined and answered separately. In order to do this, four hypotheses are used. These hypotheses are based on the literature review which was conducted and which showed what is currently known about the different aspects and what results should be expected. By doing this, all the specific elements of the research question can be answered in detail. In this way it will become clear what the different effects are of a mandatory audit partner signature on audit quality and audit fees and whether the type of industry makes a difference.
Research that has been conducted about changes in audit quality as a consequence of audit partner identification was examined in paragraph 2.1. It was shown that the studies showed mixed results: some papers showed increases in audit quality after a signature requirement was implemented, while others did not. Since this thesis conducts a study that uses data from the United Kingdom, evidence obtained with United Kingdom data is most relevant when formulating the hypotheses. The results from Carcello and Li (2013) are therefore most important. Their evidence suggests that audit quality increased as a result of the audit partner signature requirement. Because of this, the first hypothesis is:
H1: After the implementation of the audit partner’s signature requirement, the audit quality in the United Kingdom has increased.
Studies that were conducted about the effect of a mandatory audit partner signature on audit fees were discussed in paragraph 2.2. Audit fees seem to increase as a consequence of audit partner identification, although there is only limited evidence to support this effect. Only one study was conducted about this issue and therefore the results are not necessarily true. The research that was carried out, however, was undertaken in the United Kingdom. This increases its relevance for this thesis. The study by Carcello and Li (2013) showed that audit fees increased as a result of a signature requirement of the audit partner. Therefore, the second hypothesis is:
H2: After the implementation of the audit partner’s signature requirement, the audit fees in the United Kingdom have increased.
The differences in audit quality between litigious and non-litigious industries, as researched by studies, was examined in paragraph 2.3. It was found that audit quality is lower in litigious industries than in non-litigious industries. Research about the effect of audit partner
identification on changes in audit quality in both types of industries has not been conducted. Therefore, assumptions about this effect can only be made based on the difference in audit quality between the industries. It is expected that a mandatory audit partner signature would cause a bigger audit quality increase in litigious industries. The reason for this assumption is that the audit quality in these industries is more likely to improve because it is lower compared to non-litigious industries. As a result, the third hypothesis is:
H3: After the implementation of the audit partner’s signature requirement, the audit quality in the United Kingdom has increased more in litigious industries than in non-litigious industries.
Research that investigated the audit fee differences between litigious and non-litigious industries was examined in paragraph 2.4. It was clearly shown that audit fees are higher in litigious industries compared to non-litigious industries. No studies have researched the difference between both industry types in terms of audit fee changes as a consequence of an audit partner signature requirement. Because of this, only research about the difference in audit fees among both industries can be taken into account. Based on those studies, it is assumed that mandatory audit partner identification would result in a higher increase in audit fees in non-litigious industries. This is expected because audit fees are lower in these industries and are therefore more likely to rise compared to litigious industries, in which audit fees are already at a high level. Consequently, the fourth hypothesis is:
H4: After the implementation of the audit partner’s signature requirement, the audit fees in the United Kingdom have increased more in non-litigious industries than in litigious industries.
3. Methodology
3.1 Research method and sample selection
In order to answer my research question, relevant data about audit quality and audit fees will be extracted from a database and this will be analysed in order to develop conclusions on the hypothesis of this thesis. In this way, it can be determined what the effect of the mandatory audit partner signature is on audit quality and audit fees and whether or not the results differ for litigious industries in the United Kingdom. The data will be collected from DataStream. This database contains financial information for most European listed companies.
The sample that will be used consists of data from United Kingdom companies that were listed on the Financial Times Stock Exchange (FTSE) from 2008 to 2010, the years before and after the implementation of the mandatory audit partner signature. By using the data from these years, the change in audit quality and audit fees for the companies can be found by comparing the year prior to the requirement to the year after it was implemented. Firms of which there is essential data missing or firms that do not exist during all the years that are investigated will be deleted from the sample. There are some potential biases which might arise from this specific sample. For instance, because this data is from the United Kingdom, the results may not apply in the United States. If this is the case, the results of this thesis will not be useful for the PCAOB’s decision. The companies will be divided based on whether or not they operate in a litigious industry. This makes sure the differences between the industries in terms of audit quality and audit fees can be found, which enables answering the research question of this thesis. Different variables are used to measure audit quality and audit fees. Therefore, samples that will be used vary among the various variables because these are all measured in a different way.
3.2 Empirical models
In order to find useful results for this thesis, several models will be used in order to find the changes in audit quality and audit fees. The models used and the way these are measured are mainly based on the models used in the paper by Carcello and Li (2013). Audit quality is hard to determine and is therefore based on four different variables: abnormal accruals, the likelihood of firms reporting a small earnings increase, earnings informativeness and the likelihood of auditors to issue a qualified audit opinion. Audit quality has increased when firms: have lower abnormal accruals, are less likely to report a small earnings increase, have increased earnings informativeness and are more likely to issue a qualified audit opinion. In the analysis of audit fees, the model used is the natural logarithm of total audit fees.
In order to find out if there are differences between litigious and non-litigious industries in the change in audit quality and audit fees, every section contains an additional model which measures these effects. These models take into account what industry type a company operates in and what effect this has on the changes in audit quality and audit fees as a consequence of audit partner identification. The variable LITIGATE is used to show whether or not a company operates within a litigious industry. If this is the case, the variable is 1. When a company is in a non-litigious industry, the variable is 0. The following industries, which are derived from Francis, Philbrick and Schipper (1994, p. 144), are regarded as being litigious: biotechnology (SIC codes 2833-2836 & 8731-8734), computers (SIC codes 3570-3577 & 7370-7374), electronics (SIC codes 3600-3674) and retail industries (SIC codes
5200-5961). Further details about the different models and the way these are measured will now be discussed.
3.2.1 Abnormal accruals models
The effect of the audit partner signature requirement on abnormal accruals is measured by using the following model, which was used by Carcello and Li (2013, p. 1522):
ABS_ACCt = b0 + b1SIGNATURE + b2SIZEt + b3ROAt + b4LEVERAGEt + b5LOSSt + b6MBt +
b7LCACCRt + b8CFOt + b9VOLATILITYt + b10LITIGATEt + b11AUDITORt + Industry
Dummies.
ABS_ACC is measured by using abnormal accruals generated by the Modified Jones Model proposed by Dechow, Sloan and Sweeney (1995). Similar to Carcello and Li (2013, p. 1522), the model shown above controls for the size of the firm (SIZE), return on assets (ROA), whether or not there was a loss (LOSS), debt-to-assets ratio (LEVERAGE), market-to-book ratio (MB), previous year’s total current accruals (LCACCR), cash flow from operations (CFO), volatility of earnings (VOLATILITY), whether or not the firm is in a litigious industry (LITIGATE), whether or not the firm is audited by a Big 6 auditor (AUDITOR) and industry fixed effects (Industry Dummies). The expectation is that SIGNATURE is negatively related to ABS_ACC. This would mean that the audit quality would increase.
In order to discover whether or not there are differences in abnormal accruals changes between litigious and non-litigious industries, the model is modified in the following way to account for both types of industries:
ABS_ACCt = b0 + b1SIGNATURE × LITIGATE + b2SIGNATURE × NON_LITIGATE +
b3SIZEt + b4ROAt + b5LEVERAGEt + b6LOSSt + b7MBt + b8LCACCRt + b9CFOt +
b10VOLATILITYt + b11AUDITORt + Industry Dummies.
3.2.2 Small earnings increase models
The likelihood of firms reporting a small earnings increase is the second variable that is used to determine audit quality. The following model, derived from Carcello and Li (2013, p. 1522), is used to measure the effect of the mandatory signature of the audit partner on a small earnings increase:
INCREASEt = b0 + b1SIGNATURE + b2SIZEt + b3ROAt + b4LEVERAGEt + b5LOSSt + b6MBt
+ b7LCACCRt + b8CFOt + b9VOLATILITYt + b10LITIGATEt + b11AUDITORt + Industry
Dummies.
INCREASE is an indicator variable which takes the value of 1 when the earnings increase from one year to another is between zero and two per cent and takes the value of 0 otherwise. The control variables used are the same that were previously mentioned for the ABS_ACC model. It is expected that SIGNATURE is negatively associated with INCREASE, which would mean that the signature requirement would result in a lower small earnings increase and would thus result in higher audit quality.
Since this thesis also looks at the differences in changes between litigious and non-litigious industries, the model is changed to make sure it includes both industry types. In this way, the industry differences for changes in the likelihood of firms reporting a small earnings increase can be found. As a result, the following model is created:
INCREASEt = b0 + b1SIGNATURE × LITIGATE + b2SIGNATURE × NON_LITIGATE +
b3SIZEt + b4ROAt + b5LEVERAGEt + b6LOSSt + b7MBt + b8LCACCRt + b9CFOt +
b10VOLATILITYt + b11AUDITORt + Industry Dummies.
3.2.3 Earnings informativeness models
The third proxy which is used to measure audit quality is earnings informativeness, which is measured using the Earnings Response Coefficient (ERC). In order to determine the relation between the ERC and the mandatory audit partner signature, the following model of Carcello and Li (2013, p. 1523) is used:
RETt = b0 + b1ROAt + b2SIGNATURE + b3SIGNATURE × ROAt + b4SIZEt + b5SIZEt × ROAt
+ b6LEVERAGEt + b7LEVERAGEt ×ROAt + b8LOSSt + b9LOSSt × ROAt + b10MBt + b11MBt
× ROAt + b12VOLATILITYt + b13VOLATILITYt × ROAt + b14LITIGATEt + b15LITIGATEt ×
ROAt + b16AUDITORt + b17AUDITORt × ROAt + Industry Dummies.
In this model, RET represents the 12-month cumulative stock return for a certain year. The control variables that are used are the same that were used in the models which were mentioned before. SIGNATURE × ROA is expected to be positive, since it is assumed that the implementation of the audit partner signature would result in a higher association between earnings and stock returns.
The differences between litigious and non-litigious industries with respect to changes in earnings informativeness also have to be examined. This effect will be shown by using the following adjusted model, which includes both industry types:
RETt = b0 + b1ROAt + b2SIGNATURE × LITIGATE + b3SIGNATURE × NON_LITIGATE +
b4SIGNATURE × ROAt + b5SIZEt + b6SIZEt × ROAt + b7LEVERAGEt + b8LEVERAGEt ×
ROAt + b9LOSSt + b10LOSSt × ROAt + b11MBt + b12MBt × ROAt + b13VOLATILITYt +
b14VOLATILITYt × ROAt + b15AUDITORt + b16AUDITORt × ROAt + Industry Dummies.
3.2.4 Audit opinion models
Another component of audit quality is the likelihood of an auditor to issue a qualified audit opinion for a particular year. The following regression model, obtained from Carcello and Li (2013, p. 1523), is used to test the relation between the mandatory audit partner signature and the likelihood of a qualified audit opinion given by the auditor:
QUA_AOPt = b0 + b1SIGNATURE + b2SIZEt + b3ROAt + b4LEVERAGEt + b5LOSSt + b6MBt
+ b7CFOt + b8VOLATILITYt + b9LITIGATEt + b10AUDITORt + b11RECEIVABLEt +
b12INVENTORYt + b13FOREIGNt + b14BUSYt + Industry Dummies.
QUA_AOP is an indicator variable which can be either 1 or 0, depending on whether or not the auditor gives a qualified audit opinion for a certain year. The model includes the same control variables that were used in the models of the variables mentioned before, as well as four additional variables: receivables (RECEIVABLE), inventories (INVENTORY), whether or not the firm has foreign transactions (FOREIGN) and if the firm’s fiscal year-end is between December and March (BUSY). It is assumed that SIGNATURE has a positive relationship with QUA_AOP. This is because the audit partner signature requirement is expected to increase the likelihood of auditors to provide a qualified audit opinion. If this would be the case, audit quality would increase.
In order to find out if the change in the likelihood of an auditor to issue a qualified audit opinion differs between litigious and non-litigious industries, the model is modified in the following way:
QUA_AOPt = b0 + b1SIGNATURE × LITIGATE + b2SIGNATURE × NON_LITIGATE +
b3SIZEt + b4ROAt + b5LEVERAGEt + b6LOSSt + b7MBt + b8CFOt + b9VOLATILITYt +
b10AUDITORt + b11RECEIVABLEt + b12INVENTORYt + b13FOREIGNt + b14BUSYt + Industry
3.2.5 Audit fee models
This thesis also examines whether the implementation of the mandatory audit partner signature increases the amount of audit fees paid by a firm. The following model, which is derived from Carcello and Li (2013, pp. 1523-1524), is used in order to test this:
LNAFEEt = b0 + b1SIGNATURE + b2SIZEt + b3ROAt + b4LEVERAGEt + b5LOSSt + b6MBt +
b7CFOt + b8VOLATILITYt + b9LITIGATEt + b10AUDITORt + b11RECEIVABLEt +
b12INVENTORYt + b13FOREIGNt + b14BUSYt + Industry Dummies.
LNAFEE represents the natural logarithm of audit fees for a particular year. All the control variables used were also included in other models and have therefore already been explained earlier. It is assumed that SIGNATURE has a positive relationship with LNAFEE. This is because of the expectation that the mandatory audit partner signature will increase audit effort, which would lead to higher audit fees.
In order to find out what the differences are between litigious and non-litigious industries in terms of changes in audit fees, the model has to be adjusted to account for both industry types. As a result of this change, the following model is used:
LNAFEEt = b0 + b1SIGNATURE × LITIGATE + b2SIGNATURE × NON_LITIGATE + b3SIZEt
+ b4ROAt + b5LEVERAGEt + b6LOSSt + b7MBt + b8CFOt + b9VOLATILITYt + b10AUDITORt +
b11RECEIVABLEt + b12INVENTORYt + b13FOREIGNt + b14BUSYt + Industry Dummies.
4. Empirical Results
This section contains the empirical results which have been generated in order to find out whether or not the audit partner signature requirement influences the audit quality and audit fees. Also, the differences between the consequences on litigious and non-litigious industries in terms of audit quality and audit fees changes will be shown. The data collected from DataStream was used in order to derive the results. Some observations with extreme values were excluded. Using the models which were explained in the previous section, multiple tests were performed using Stata to generate the output. Summary statistics, correlation tables and regression outputs were produced and are shown below in order to be able to answer the research question of this thesis. For each model, the different results will be discussed to show what the effect of the audit partner signature is on the particular variable.
Table 1A: Summary Statistics for Dependent Variables in 2008
Table 1B: Summary Statistics for Dependent Variables in 2010
Table 2: Correlation Coefficients (and p-Values) between Dependent Variables and SIGNATURE Variables
Table 3A: Coefficients (and p-Values) of Regression Output and Model Statistics Variable Number of
Observations Mean Value
Standard Deviation Minimum Value Maximum Value Median ABS_ACC 391 0.428 0.397 -1.381 2.493 0.382 INCREASE 515 0.014 0.116 0.000 1.000 0.000 RET 541 0.145 0.432 0.001 4.842 0.007 QUA_AOP 544 0.015 0.120 0.000 1.000 0.000 LNAFEE 517 6.014 2.021 2.303 14.827 6.215 Variable Number of
Observations Mean Value
Standard Deviation Minimum Value Maximum Value Median ABS_ACC 391 0.371 0.390 -0.833 2.890 0.277 INCREASE 515 0.021 0.145 0.000 1.000 0.000 RET 541 0.141 0.505 0.001 7.134 0.006 QUA_AOP 544 0.024 0.153 0.000 1.000 0.000 LNAFEE 517 6.083 1.913 2.485 10.933 6.216
Variable ABS_ACC INCREASE RET QUA_AOP LNAFEE
SIGNATURE -0.134 (0.005) -0.087 (0.102) -0.038 (0.579) -0.075 (0.157) 0.180 (0.004) SIGNATURE×LITIGATE 0.126 (0.008) 0.058 (0.275) 0.203 (0.003) 0.588 (0.000) -0.099 (0.118)
Variable ABS_ACC INCREASE RET QUA_AOP LNAFEE
SIGNATURE -0.066 (0.006) -0.656 (0.092) -0.253 (0.008) -0.885 (0.041) 0.579 (0.007) SIZE 0.016 (0.281) -0.173 (0.478) 0.322 (0.000) -0.446 (0.014) 0.215 (0.036) ROA 0.187 (0.082) 5.118 (0.256) 1.584 (0.001) 10.748 (0.082) 3.206 (0.003) LEVERAGE 0.081 (0.304) -3.329 (0.183) -0.134 (0.736) -1.742 (0.394) 1.823 (0.252) LOSS -0.005 (0.872) 1.000 (0.032) -0.251 (0.021) 0.413 (0.319) 0.091 (0.714) MB -0.001 (0.786) -1.538 (0.051) -0.004 (0.715) 0.018 (0.897) -0.007 (0.671) LCACCR 0.880 (0.041) -2.688 (0.109) CFO -0.007 (0.495) 9.614 (0.048) -0.513 (0.298) -0.276 (0.135) VOLATILITY 0.000 (0.156) 0.390 (0.029) 0.000 (0.077) 0.145 (0.037) -0.114 (0.064) LITIGATE 0.071 (0.025) 0.114 (0.820) 0.403 (0.001) -0.731 (0.259) -0.413 (0.093) AUDITOR 0.509 (0.001) -3.021 (0.000) -0.220 (0.555) -3.243 (0.007) -0.491 (0.670) RECEIVABLE -5.529 (0.117) 5.308 (0.000) INVENTORY -2.374 (0.119) 3.708 (0.082) FOREIGN -0.031 (0.937) 0.629 (0.003) BUSY 0.283 (0.481) -0.059 (0.787) Intercept -0.395 (0.137) 3.971 (0.256) -3.568 (0.000) 6.972 (0.016) 2.423 (0.167) Model Statistic N 782 1030 1082 1088 1034 (Pseudo) R² 0.083 0.392 0.360 0.396 0.295 Prob>F/Chi² 0.000 0.000 0.000 0.001 0.000
Table 3B: Coefficients (and p-Values) of Regression Output and Model Statistics
4.1 Abnormal accruals results
As was mentioned in the methodology section, abnormal accruals is one of the components which is used to measure audit quality. Beforehand, it was expected that the variable SIGNATURE is negatively related to ABS_ACC. This would mean that the audit quality would increase after the implementation of the audit partner signature requirement. In Tables 1A and 1B, the summary statistics for ABS_ACC are shown. These tables include the mean values for this variable for both 2008 and 2010, the years before and after the audit partner identification requirement. The mean value of ABS_ACC decreased from 0.428 to 0.371, indicating that the audit partner signature requirement reduced the amount of abnormal accruals. Table 2 shows the correlations between the SIGNATURE variables and the dependent variables. The model used to calculate abnormal accruals contains the variable SIGNATURE, which indicates whether or not a certain year contains the signature requirement. As can be seen in the table, ABS_ACC and SIGNATURE are negatively correlated with a value of -0.134. This shows that the identification of the auditor decreases the abnormal accruals. The outputs of the regressions are shown in Table 3A. The R² of the model is quite low at 0.083, but this is consistent with prior literature. The results show the association between the audit partner signature requirement and the amount of abnormal
Variable ABS_ACC INCREASE RET QUA_AOP LNAFEE
SIGNATURE×LITIGATE ×LITIGATE 0.094 (0.009) 0.957 (0.079) 0.350 (0.007) 2.177 (0.000) -0.649 (0.020) SIZE 0.014 (0.354) -0.164 (0.517) 0.321 (0.000) -0.358 (0.062) 0.200 (0.052) ROA 0.138 (0.196) 7.100 (0.183) 1.444 (0.003) 9.326 (0.198) 3.621 (0.001) LEVERAGE 0.088 (0.264) -3.476 (0.168) -0.112 (0.780) -1.237 (0.603) 2.060 (0.199) LOSS 0.010 (0.711) 1.394 (0.010) -0.157 (0.142) 0.090 (0.860) -0.035 (0.890) MB -0.001 (0.824) -1.463 (0.061) -0.003 (0.829) 0.227 (0.418) -0.007 (0.684) LCACCR 1.045 (0.015) -2.579 (0.139) CFO -0.006 (0.543) 9.917 (0.045) -1.273 (0.490) -0.304 (0.102) VOLATILITY -0.018 (0.116) 0.473 (0.015) 0.000 (0.062) 0.093 (0.196) -0.125 (0.044) LITIGATE AUDITOR 0.478 (0.002) -3.450 (0.000) -0.184 (0.631) -2.368 (0.476) -0.365 (0.753) RECEIVABLE -3.925 (0.320) 5.322 (0.000) INVENTORY -2.100 (0.335) 4.020 (0.060) FOREIGN 0.308 (0.490) 0.609 (0.005) BUSY 0.260 (0.576) -0.045 (0.836) Intercept -0.376 (0.159) 3.284 (0.354) -3.747 (0.000) 3.962 (0.352) 2.884 (0.102) Model Statistic N 782 1030 1082 1088 1034 (Pseudo) R² 0.070 0.394 0.323 0.512 0.280 Prob>F/Chi² 0.000 0.000 0.000 0.000 0.000
accruals. The variable SIGNATURE is significantly negative (p=0.006) with a value of -0.066. This result means that the audit partner identification reduces the abnormal accruals. All other control variables are in accordance with expectations. Overall, the results of the various tests which were performed indicate that the implementation of an audit partner signature requirement decreased the amount of abnormal accruals. This is in line with the expectations based on prior literature and shows that, solely based on abnormal accruals, the audit quality increased after an audit partner identification requirement.
In addition to the overall change in abnormal accruals, the differences between the changes for companies in litigious and non-litigious industries are also examined. This is done to find out whether or not the change in audit quality, as a consequence of the audit partner signature, differs between both industry types. The expectation is that the audit quality increase is higher in litigious industries than in non-litigious industries. Since abnormal accruals is a component of audit quality, this would mean that the decrease of this variable would be higher for companies that operate in a litigious industry. In Table 2, it is shown that the abnormal accruals model contains the variable SIGNATURE×LITIGATE. This variable represents the number of companies operating in a litigious industry in the period following the signature requirement. The correlation between SIGNATURE×LITIGATE and ABS_ACC is 0.126. This indicates that the audit partner signature requirement increased abnormal accruals in litigious industries. Table 3B shows the regression results for the abnormal accruals model, which indicates the relation between audit partner identification and the amount of abnormal accruals in litigious industries. For this model, the R² is 0.070. Although this seems rather low, it is similar to models from past literature. The variable SIGNATURE×LITIGATE is significantly positive (p=0.009) with a value of 0.094, showing that abnormal accruals increased for companies within litigious industries as a consequence of the audit partner signature requirement. The other control variables are consistent with the values that were expected in advance. The abnormal accruals as a whole decreased following the introduction of the signature requirement. The results from both the correlation tables and regression outputs show that abnormal accruals increased in litigious industries following the engagement partner signature requirement. Because of these results, it seems clear that the abnormal accruals for companies within non-litigious industries decreased more than for companies within litigious industries. This result differs from the anticipated outcome, since it suggests that audit quality increased more in non-litigious industries compared to litigious industries.
