Measures for enhancing auditor independence: Perceptions of spanish non-professional investors and auditors

Abstract

Auditors fulfill a crucial societal role by increasing the credibility of financial reports, which requires that appropriate audit quality be provided and auditors are factually independent and perceived as such. So far, regulators and researchers have focused on a few measures for improving independence, namely the prohibition of non-audit services (NAS), audit firm and audit partner rotation, and joint audits. Despite these regulatory initiatives, accounting scandals still happen, and the independence of the auditors involved is questioned. Therefore, this explorative study aims to identify alternative measures that may strengthen auditor independence. Non-professional investors and auditors participated in a survey to identify promising independence-enhancing measures.

Surprisingly, not only non-professional investors but also auditors support the implementation of many instruments. Non-professional investors rate measures related to oversight, controls, and sanctions as particularly useful. Likewise, auditors perceive measures for oversight and control as helpful but rank most sanctions lower than non-professional investors. They also support a stricter civil liability and rank some direct independence-improving measures as highly useful. Overall, the participants perceive a greater need for improving the independence of public interest entity (PIE) auditors.

This explorative study widens the research horizon by adding alternative measures for improving independence in appearance. Thus, researchers may identify promising avenues for future research. Also, the paper is of interest to regulators and potentially helps in setting independence standards. Auditors and audit committees can identify opportunities for taking voluntary actions to ensure factual and perceived auditor independence.

1

Introduction

Auditors are entrusted by law to conduct statutory audits. They should provide reasonable assurance as to whether the financial statements are prepared according to the applicable financial reporting framework and are free from material misstatements (ISA 200.5). Audits fulfill an important societal role in providing an opinion on the truth and fairness of audited entities' financial statements, reducing the risk of misstatement, and enhancing the degree of confidence of intended users in financial reports prepared by management (Watts & Zimmermann, 1986). Audits are thus a crucial contributor to financial stability, trust, and market confidence in the economy by protecting investors from agency risk and thereby reducing the cost of capital for companies (Jesover & Kirkpatrick, 2005; European Commission, 2010). For the fulfillment of this role, appropriate audit quality is fundamental. According to the generally accepted definition of DeAngelo (1981), audit quality is the market-assessed joint probability that an auditor will discover material misstatements (auditor competence) and also report them (auditor independence). This definition stresses that providing a high factual audit quality is insufficient, but users must also perceive audit quality as appropriate. The independence of auditors is thus the fundamental bedrock of the audit environment. Regulators need to consistently reinforce auditor independence and address conflicts of interest to ensure that audits are fit for the purpose (European Commission, 2010).

The International Ethics Standards Board for Accountants (IESBA) of the International Federation of Accountants distinguishes between independence of mind and independence in appearance (IESBA 2021). The former is defined as ‘the state of mind that permits the expression of a conclusion without being affected by influences that compromise professional judgment, thereby allowing an individual to act with integrity and exercise objectivity and professional skepticism’ (IESBA 2021, Sec. 400.5 (a)). Independence in appearance is defined as ‘the avoidance of facts and circumstances that are so significant that a reasonable and informed third party would be likely to conclude that a firm's, or an audit team member's, integrity, objectivity or professional skepticism has been compromised’ (IESBA 2021, Sec. 400.5 (b).

Threats to the fundamental principle of independence might be created by a broad range of facts and circumstances which fall into one or more of the following categories (IESBA 2021, Sec. 120.6 A3): (1) Self-interest threat – the threat that a financial or other interest will inappropriately influence auditor judgement or behavior (economic bonding); (2) Self-review threat – the threat that the auditor will not appropriately evaluate the results of a previous judgment made, or an activity performed by him or another individual from the same audit firm; (3) Advocacy threat – the threat that an auditor will promote a client's position to the point that the auditor's independence is compromised; (4) Familiarity threat – the threat that due to a long or close relationship with the client, an auditor will be too sympathetic to the client's interests (social bonding); (5) Intimidation threat – the threat that an auditor will be deterred from acting independently because of actual or perceived pressures. Measures for enhancing auditor independence must combat such threats.

Auditor independence has been debated for decades, and after several corporate failures, it is in the news again (Ketz, 2020). Recent regulatory initiatives focused mainly on auditor tenure and the provision of NAS. In 2016, e.g., the EU regulation on statutory audits of PIE came into force. It introduced, among others, a blacklist of prohibited NAS, a non-audit service fee cap, and mandatory audit firm rotation after ten years. Despite such regulatory changes, major accounting scandals occurred, e.g., Wirecard in Germany, Carillion in the UK, and Pescanova and Grupo Dia in Spain. Thus, auditor independence remains controversial, indicating that these attempts to improve auditor independence are inappropriate or at least insufficient (Hilzenrath, 2022). Prior research also primarily addressed the relationship between non-audit service fees and factual or perceived audit quality, as well as the quality impacts of audit tenure or auditor rotation. Therefore, researchers called for research on alternative measures to promote auditor independence (Church et al., 2015). Thus, it is important to identify additional or alternative measures to strengthen auditor independence, and thereby also promising avenues for future research.

The primary purpose of this study is to identify measures that non-professional investors and auditors perceive as beneficial for enhancing auditor independence. To achieve this objective, a survey was performed. The questionnaire used encompasses 50 potentially useful measures developed from the literature and expert interviews. The participants in the survey were non-professional investors and auditors from Spain. The information gap between management and non-professional investors is high, and thus, they are particularly vulnerable. Their answers reflect impacts on independence perceptions. In contrast, auditors are highly knowledgeable, and their responses may indicate effects on factual independence. Spain is a typical Continental European country with an environmental setting that substantially differs from those of Anglo-Saxon countries and is characterized, e.g., by a Roman law system, a moderate level of investor protection (La Porta et al., 1997), and the high relevance of banks to the financial system (Carrasco et al., 2014).

Results show that non-professional investors mainly perceive indirect measures, such as oversight, controls, and sanctions, as highly suitable for improving auditor independence. By contrast, the preferences of auditors are more mixed and also include direct measures such as mandatory rotation of audit team members and further prohibitions of auditor-provided NAS. In general, auditors assess sanctions as less suitable than non-professional investors; however, they also rank a stricter civil liability of the auditor relatively high. Both subject groups suggest a more substantial need for enhanced auditor independence for PIE audits than non-PIE audits, as expressed by higher means, i.e., higher suitability regarding all proposed measures and a higher number of measures significantly perceived as improving independence. Furthermore, auditors see a stronger need for strengthening auditor independence regarding PIE audits than non-professional investors. By contrast, the demands from non-professional investors for an improvement in auditor independence regarding non-PIE audits exceeds that of auditors. Finally, the demographic characteristics of the participants are significantly associated with their evaluation of the independence-enhancing measures.

Prior research on measures for improving auditor independence mainly have focused on restrictions regarding the provision of NAS and mandatory audit firm and audit partner rotation, and other instruments are barely considered. This paper substantially widens the research horizon by adding promising measures for improving independence in appearance. Due to the explorative nature of the study, the results should be beneficial for researchers searching for promising new avenues for research concerning auditor independence. Its findings are also of interest to regulators who might consider implementing additional or alternative measures for enhancing auditor independence. Auditors and audit clients may identify opportunities for voluntary and proactive actions to improve independence and, thereby, the audit quality perceptions of investors.

The remainder of the paper is organized as follows. Section 2 provides an overview of the Spanish auditing setting. Section 3 includes background information, describes prior research on the provision of NAS, audit firm rotation, and joint audits, and formulates the research questions, whereas Section 4 describes the research method. The study results are shown and discussed in Section 5. After discussion included in the Section 6, the final Section 7 concludes the article.

2

Spanish environmental setting

2.1

Statutory audits and engagement of the auditor

Spanish law requires statutory audits for medium-sized and large corporations, and cooperatives. Small corporations and cooperatives are exempted from statutory audits².

Above and beyond this, audits are mandatory for all PIE, i.e., capital market-oriented companies, banks, insurance companies, investment companies serving more than 5000 clients or with more than 5000 members in two consecutive years, pension funds serving more than 10,000 clients in two consecutive years, trust companies in banking, payment service providers and unlimited companies exceeding sales of two billion Euros and with more than 4000 employees in two consecutive years (Article 15 of Real Decreto 877/2015, de desarrollo de la Ley 26/2013 de 27 de diciembre de cajas de ahorros y fundaciones bancarias). Finally, audits are mandatory for companies that have received public subsidies or aids, for companies that received public orders for more than 600,000 Euro if they exceed 50 % of total sales, large foundations, large housing cooperatives, and cooperative banks.

Principally, the administration board suggests, and the board of directors engages the auditor (Article 264 LSC). In case that an audit committee exists, it recommends an auditor in accordance to Article 16,2 of the EU Regulation. If the auditor engagement is not renewed, the audit committee must submit at least two suggestions.

The auditor will be hired initially for not less than three years and not more than nine years and may be hired for successive maximum periods of up to three years once the initial period has ended (Article 22 Ley 22/2015, de 20 de julio, de Auditoría de Cuentas, LAC). For audits of public-interest entities, Article 17 of the EU Regulation applies. It requires an external rotation of the auditor after ten years and audit partner rotation after seven years. The Regulation also included a Member State option to extend the maximum duration of an audit engagement to 20 years when a pubic tendering process for the statutory audit is conducted and to 24 years when more than one auditor is simultaneously engaged. Spain made use of this option (Article 40,1 LAC). The Regulation also allowed Member States to require an earlier audit partner rotation. Thus, Spanish law says that the key audit partner responsible for carrying out a statutory audit shall cease participation after five years (Article 40,2 LAC).

2.2

Auditor independence

Auditors must be independent of their clients and refrain from participating in the audited entity's management or decision-making (Article 14 LAC). Labor, commercial, or other relationships with the audited company may generate a conflict of interest or be perceived as such.

Articles 16 and 18 LAC list situations in which it is incontestably assumed that the auditor or audit firms lack independence:

• •

  Being a board member, a manager, an employee in financial accounting, or in charge of the internal control system of the client,

• •

  Having a financial interest in the client (e.g., being a shareholder),

• •

  Provision of financial services,

• •

  Assisting in bookkeeping or the preparation of financial statements,

• •

  Provision of valuation services with a material impact on the financial statements to be audited,

• •

  Provision of internal audit services,

• •

  Provision of legal services,

• •

  Design and implementation of internal control systems,

• •

  Family relationships with board members or with managers in the accounting department,

• •

  Financial interests of the auditor's spouse in the client and

• •

  Family members of the auditor take part in the emission of client securities.

Moreover, the regulation excludes an auditor or an audit firm if the fees from one client exceed 30 % of the average total revenue of the last three years (Article 64 Reglamento de la Ley de Auditoría RD 2/2021, RLA).

2.3

Public oversight

The Instituto de Contabilidad y Auditoría de Cuentas (ICAC) is the public audit oversight entity in Spain. ICAC's responsibilities include authorizing auditors and audit firms to be registered in the official registry of auditors and maintaining this registry, monitoring and overseeing auditors’ fulfillment of requirements, and establishing and operating a quality assurance review system for audits (Article 46 LAC). The supervision of auditors is carried out by investigations and inspections (Article 52 LAC), and both are performed by ICAC staff. However, it is possible to hire third parties if there is a lack of capacity or expertise (Article 55 LAC).

Investigations aim to determine facts or circumstances that may indicate possible breaches of audit regulations. They include examining the audit work files or other documentation, conducting inquiries, and obtaining and evaluating any additional relevant information or documentation (Article 53 LAC). The inspections are regulated by Article 54 LAC and Articles 100-104 RLA and performed by employees of the ICAC. They consist of the periodic review of the internal quality control systems of registered auditors and audit firms. The selection of auditors, as well as of the elements of the quality control system and the files of the audit work to be reviewed, are based on the risk criteria of the auditors and the audited entities. Inspections take place at least every third year. However, auditors who only audit non-PIE are subject to inspections at least every six years. The inspections include at least the verification of the internal quality control system of the auditors, the review of the procedures documented in the audit files to verify the proper implementation of this system, as well as, where appropriate, the monitoring of the improvement measures implemented to correct the deficiencies revealed in the immediately preceding inspection. The results of the inspections are documented in a report. If the report refers to PIE auditors, it will be on the ICAC website without identifying the audited entities' data.

In cases of revealed misconduct, the ICAC will sanction the auditor or the audit firm. For this purpose, Article LAC classifies the committed infractions as very serious, serious, or minor (Article 71 LAC). Intentional and grossly negligent violations of the duty of independence are considered very serious infractions (Article 72 LAC). Otherwise, the infraction is classified as serious.

2.4

Disciplinary sanctions

Article 75 LAC regulates sanctions for infractions committed by an individual auditor. Prior literature has discussed these sanctions (Castellanos Rufo & Gonzalo Angulo, 2010; García Benau & Maestro Chillida, 2012). One of the following sanctions can be imposed on the offender: (a) Revocation of authorization and removal from the register of auditors; (b) Suspension of authorization and temporal withdrawal from the register of auditors for a period of more than two and up to five years; (c) Fine amounting to six to nine times the fee for audit work in which the infraction was committed; the fine should not be less than 18,001 and not more than 36,000 Euros; this range also applies when the infraction has not been committed in conjunction with specific audit work. For serious infractions, the possible sanctions are (a) Suspension of up to two years; (b) Fine amounting to two up to five times the fee for audit work in which the infraction was committed; with a range from 6001 to 18,000 Euros.

Regarding infractions committed by audit firms, Article 76 LAC mentions two sanctions for very serious infractions: (a) Withdrawal of authorization and removal from the register of auditors; (b) Fine of between three and six percent of the audit firm's total fees billed for audits in the previous year, but at minimum 24,000 Euros. The engagement partner can be imposed upon by: (a) Withdrawal of authorization and removal from the register of auditors; (b) Suspension for a period of two years and one day to five years; (c) Fine of a minimum amount of 12,001 Euros and a maximum of 24,000 Euros. For serious infractions, the offending audit firm will be sanctioned with a fine of up to three percent of the audit firm's total fees billed for audits in the previous year, but at minimum 12,000 Euros. The engagement partner may be suspended for up to two years or a fine with a minimum of 3,000 Euros and maximum of 12,000 Euros.

In the case of PIE audits, the amount of a fine may be increased up to 20 %. The minimum and maximum amounts will be increased accordingly. Furthermore, in addition to a fine, the audit firm and the engagement partner may be suspended from carrying out PIE audits for up to two years in cases of serious infractions, and five years in cases of very serious infractions Article 77 LAC).

When the imposition of a sanction for a very serious or serious infraction is a consequence of an audit of a specific client, this sanction entails the prohibition of the individual auditor or the audit firm and the engagement partner from auditing the same entity again in the first three fiscal years that begin after the date on which the sanction becomes firm (Article 78 LAC).

All sanctioned infractions, except private reprimands imposed for a minor offense, will be published in the Bulletin of the ICAC (Article 82 LAC). The official State Gazette publishes sanctions for infractions in conjunction with PIE audits. The publication includes information on the type and nature of the infraction and the identity of sanctioned auditors and audit firms.

2.5

Civil liability

According to Article 26 LAC, auditors are proportionally liable (Humphrey et al., 2003) to the audited entity and third parties for damages arising from their misconduct, and the general rules of the Civil Code (CC) apply (Velasco Fabra, 2013; for the historical development, see Ruiz Barbadillo et al., 2000). Thus, liability requires the fulfillment of four prerequisites, an action by the auditor contrary to the regulations; fault, i.e., intention or negligence on the part of the auditor; damage must be caused to the company or a third party; and a causal relationship between the action carried out by the auditor and the damage cause (Velasco Fabra, 2013).

The civil liability of auditors can be of two types, contractual (liability against the client) and non-contractual (liability against third parties with whom the auditor has no contractual relationship). The debtor, i.e., the auditor, must compensate the creditor, i.e., the client, for damages incurred through fraud, negligence, or delay in fulfilling obligations (Articles 1101 CC). Above and beyond that, a third party could have considered the auditor's report as an essential element for motivating action or decision-making because of this report's reliability. Regarding third-party liability, Article 1902 CC is pertinent, stating that the party who, by action or omission, intentionally or negligently causes damage to another is obliged to compensate for the damage incurred.

Generally, the damage must be compensated for whenever the prerequisite conditions are met. This does not, however, prevent flexible judicial decision-making. Thus, not every act of the defendant will constitute a fault, and not all harm claimed will be worth compensating. The Spanish courts do not hesitate to employ the above conditions to restrict liability and allow themselves plenty of leeway to avoid having to order compensation for excessively remote damages (Abdussamad, 2015). According to Articles 1101 and 1902 CC, auditors shall only respond to losses that are necessary consequences of their personal faults. Auditors cannot be burdened with the fault of management. Moreover, real damage is required, and the probability of damage is insufficient (Muñoz Delgado, 2018). Damage caused by the audited entity or by third parties is excluded. It is up to the plaintiff, following the rules on the burden of proof contained in article 217 of Law 1/2000, of January 7, on Civil Procedure, to prove that it was the act of the auditor that caused the harmful result for which compensation is claimed (Artículo 217 de la Ley 1/2000, de 7 de enero, de Enjuiciamiento Civil).

The action of demanding the contractual liability of the auditor becomes time-barred after four years from the date of the audit report (Article 26,4 LAC). Non-contractual liability claimed by third parties who relied on the auditor's report is subject to the general one-year limitation period (Article 1968 CC). When an auditor audits accounts on behalf of an audit firm, the auditor who signed the audit report and the audit firm are jointly and severally liable. It is not possible to limit auditor liability by contract (Artículo 11,2 Real Decreto 2/2021, de 12 de enero, por el que se aprueba el Reglamento de desarrollo de la Ley 22/2015, de 20 de julio, de Auditoría de Cuentas).

3

Background, prior literature, and research questions

3.1

Background

There is a principal-agent relationship between the management of a firm and its owners (Jensen & Meckling, 1976). Information asymmetries characterize this relationship; that is, the management has more, better, and earlier information. Consequently, shareholders cannot entirely observe and assess management's performance (hidden action; Arrow, 1985), and the management could be tempted to use its advanced information for maximizing its own benefit instead of that of the stakeholders (moral hazard; Hart & Moore, 1990). Such deviations from the principal's interest are called agency costs (Jensen, 1986). Disclosing financial statements may reduce information asymmetries (Watts & Zimmerman, 1978). However, financial statements are prepared by management; thus, there is again a hidden action situation and a moral hazard risk. For reducing related agency problems, external auditors are needed to attest to the purported accuracy and completeness of company financial information (Wallace, 2004) and serve as an objective and trustworthy arbiter of the fair presentation of financial statements (Falk et al., 1999). The audit opinion serves as a signal to investors about the reliability and credibility of financial statement information (Healy & Palepu, 2001).

The audit function can only be fulfilled if appropriate quality is provided. However, audits are credence goods. The outcome of an audit is unobservable, and the idiosyncratic and uncertain nature of the audit process means that only the auditor can decide how much effort to exert and what evidence to gather (Causholli & Knechel, 2012). Again, information asymmetries between a potential auditor and the shareholders regarding the auditor's attributes exist, and such hidden characteristics (Rothschild & Stiglitz, 1978) could result in an adverse selection problem (Akerlof, 1978). Moreover, there is a hidden action problem (Laffont & Tirole, 1993) because the audit effort remains unclear, causing a moral hazard risk (Jensen & Meckling, 1976); that is, the assurance provider could provide a lower audit quality than the client and its shareholders assume. Audit quality depends not only on auditor competence but also on auditor independence (DeAngelo, 1981). Signaling theory is defined as the deliberate reduction of information asymmetries in a market through the provision of information through a costly signal by one party (Spence, 1973). Implementing independence-enhancing measures by regulators or their voluntary application by audit firms could serve as such a signal.

The concept of independence is one of the cornerstones of the auditing profession and an essential ingredient of user confidence in financial reports (Mautz & Sharaf, 1961, 246). Consequently, regulators and oversight bodies require the auditor to be independent of the client (Church et al., 2018). The IESBA recognizes two types of auditor independence, independence of mind and independence in appearance. The former reflects an auditor's state of mind that allows the audit to be performed with an unbiased attitude (Campa & Donnelly, 2016). The latter refers to the avoidance of facts and circumstances that would lead users of the auditor's report to conclude that the auditor would be unable to act without bias (Dopuch et al., 2003).

Unfortunately, the auditor's mindset is unobservable. Thus, it is difficult to determine whether the auditor is independent of mind. Independence in appearance is associated with the nature of the auditor-client relationship and entails whether others believe that the auditor is impartial and free of conflicts of interest (Church et al., 2018). Regulators have focused their rulemaking almost exclusively on independence in appearance. Consequently, most rules are related to the prohibition of relationships between auditor and client (Fearnley et al., 2005). In particular, the association between fees received by audit firms directly from clients (Blay & Geiger, 2013), the economic dependence resulting from the provision of NAS by audit firms, and the familiarity developed from lengthy auditor tenure (Hussey, 1999) were alleged to contribute to this erosion of auditor independence. To restore public confidence, regulators implemented policies such as mandatory audit firm and audit partner rotation and the prohibition of certain types of NAS (Ye et al., 2011).

In 2014, in response to the global financial crisis, the EU passed its regulation on statutory audits of PIE which came into force two years later. It prohibits contingent audit fees and caps the total fees received from one PIE to 15 % of the total revenue of the audit firm. When the firm exceeds this threshold in three consecutive financial years, it must disclose this to the audit committee and discuss the threats to auditor independence and the safeguards applied to mitigate those threats, with the audit committee. Where these fees continue to exceed the 15 % threshold, the statutory audit can only be continued for two years (Article 4, 3.). Moreover, the regulation includes an extensive list of prohibited NAS (Article 5, 1.), and it caps NAS fees from audit clients to 70 % of the average audit fees from the last three years (Article 4, 2.). Finally, the regulation introduced mandatory audit firm rotation after ten years (Article 17,1.) and confirmed compulsory audit partner rotation after seven years (Article 17, 7.).

In the UK, accounting scandals, e.g., at BHS, Carillion, or Thomas Cook, triggered regulatory responses, mainly focusing on expanding the prohibition of NAS to audit clients (Department for Business, Energy & Industrial Strategy, 2021). The provision of NAS to a PIE audit client is now limited to services regarding mandatory reports and audit-related services (FRC, 2019). Furthermore, the Big 4 audit firms have to operationally separate their audit and NAS practices by 30 June 2024 (FRC, 2020). The UK regulator even considered an audit-only firm approach (Marriage, 2018). Finally, the UK government plans a managed shared audit regime, i.e., it will require UK-incorporated FTSE 350 companies to appoint a challenger as the sole group auditor or appoint a challenger firm to conduct a sufficient proportion of its subsidiary audits within a shared audit (Department for Business, Energy & Industrial Strategy, 2022). In shared audits, one firm is appointed to audit the group accounts and, usually, some of the components, while another firm or firms, audits the other components, representing a significant share of the total group (ICAEW, 2022).

After the Wirecard accounting scandal resulting in the insolvency of this German payment processor and financial service provider, the German legislator took measures to improve auditor independence and audit quality in 2021 (for an intensive discussion, see Quick, 2021a). Concerning statutory audits, certain tax and valuation services were prohibited, and German use of the member state option to extend the maximum engagement duration to 20 years in case of a public tendering process, and 24 years in cases of joint audits (Article 17, 4. EU Regulation), were reverted. Additionally, Germany tightened the audit partner rotation cycle to five years (§ 43, 6 WPO = Wirtschaftsprüferordnung = Public Accountant Act). Finally, the auditor civil liability caps were increased significantly, and the law now differentiates between simple and gross negligence (§ 323, 2 HGB = Handelsgesetzbuch = Commercial Code). The European Commission also considers increasing liability caps or eliminating them (Quick, 2021b).

The recent regulatory actions have been exclusively focused on audits of PIEs, potentially indicating a lack of perceive need for strengthening the independence of non-PIEs. These assumptions require verification.

3.2

Prior research

Like regulators, researchers have focused on just a few topics related to auditor independence. There are mainly three potentially independence-enhancing measures addressed by prior research: the (non-) provision of NAS, audit firm rotation, and joint audits.

3.2.1

The provision of NAS

The simultaneous provision of audit and NAS may result in economic and social bonding and thereby threaten auditor independence (Svanström, 2013). NAS have higher profit margins than audit services, and total revenue from one particular client increase as a result, creating an economic bond between auditor and auditee (self-interest threat, Arruñada, 1999). Furthermore, the NAS provision is based on a specific bond of trust between the consultant (i.e., the audit firm) and client management. Because of such a close relationship (social bonding), the auditor may be too sympathetic to the client's interests (familiarity threat). Another reason for the potentially harmful effect of high NAS fee levels is that focusing on NAS provision could distract and detract from auditing services (Beardsley et al., 2019). However, the provision of NAS to audit clients could enhance the auditor's ability to detect material misstatements through knowledge spillovers (Knechel et al., 2012).

Apart from a few experiments (e.g., Kowaleski et al., 2018), prior research on the impact of NAS fees on factual audit quality is archival.³ Archival studies only observe the net effect of a joint provision of audit and NAS on audit quality, i.e., the difference between a potentially positive impact on auditor competence and a potentially negative impact on auditor independence. Audits are credence goods; thus, audit quality is largely unobservable (Causholli & Knechel, 2012). Consequently, archival research must apply audit quality proxies to assess the impact of auditor-provided NAS on audit quality, and shows somewhat mixed results. However, most prior studies failed to identify a significant relationship between NAS fees and audit quality. Some studies even showed a positive impact, in particular for tax services (e.g., Watrin et al., 2019; Castillo-Merino et al., 2020). The most frequently applied audit quality proxies are modified audit opinions (e.g., Ianniello, 2012), going concern opinions (e.g., Lennox, 2016), the level of earnings management (e.g., Garcia-Blandon et al., 2020; Donelson et al., 2020, Beardsley et al., 2021), and restatements (e.g., Lisic et al., 2019; Shi et al., 2023).

Prior research analyzing the impact of NAS on perceived auditor independence or perceived audit quality depicts a different pattern and frequently reveals a negative effect. Applied research methods are archival studies, which often measure capital market reactions (earnings response coefficients, abnormal return, the value relevance of earnings, cost of capital, yield spreads, the market value of equity; e.g., Alsadoun et al., 2018; Chen et al., 2019; Alrashidi et al., 2021), credit risk assessments (Zalata et al., 2020), or auditor ratification votes (e.g., Mishra et al., 2005) to proxy audit quality, experiments (e.g., Aschauer & Quick, 2018; Meuwissen & Quick, 2019), and surveys (e.g., van Liempd et al., 2019; García-Hernández et al., 2022). Again, results regarding tax fees yield a different pattern, often with positive perceptions (e.g., Eilifsen et al., 2018).

Research using data from years after the EU regulation also demonstrates the negative impacts of NAS fees (e.g., Castillo-Merino et al., 2020; Friedrich et al., 2022).

The relationship between NAS fees and perceptions of auditor independence or audit quality differs by type of service, confirming the EU blacklist approach. A general prohibition of the joint provision of audit and NAS, or even an audit firm only-approach, seems unnecessarily strict. Moreover, the perceptions vary between stakeholder groups. The lower the level of auditing expertise of subjects, the more negative their independence perceptions. Thus, regulators face the difficult decision as to which stakeholders they should address prohibitions of NAS.

3.2.2

Mandatory rotation of audit firms

Audit firm rotation reduces clients’ opportunities to pressure the auditor by threatening to discontinue the engagement because it limits economic rents that the auditor can obtain from a given client (Arrunada & Paz-Ares, 1997; Jennings et al., 2006). In addition, it could prevent social bonds between the auditor and the client firm, which could arise from long auditor-client relationships (Boone et al., 2008; Patterson et al., 2019). Thus, rotation may strengthen auditor independence. However, it also leads to a loss of client-specific knowledge, reducing the auditor's ability to detect material misstatements (Johnson et al., 2002). Thus, the overall effect on audit quality is theoretically unclear.

Prior archival auditing research about the effect of audit firm rotation on factual audit quality relies on the few settings that have mandated auditor rotation before the EU regulation. Evidence from South Korea is mixed, with Kim & Yi (2009) demonstrating improved audit quality after rotation, Kwon et al. (2014) finding no significant effect, and Mali & Lim (2018) revealing a negative impact. Similarly, Corbella et al. (2015) find positive effects of rotation on audit quality in the Italian setting, but only for non-Big 4 clients. In contrast, studies by Cameran et al. (2015) and Cameran et al. (2016) demonstrate reduced audit quality. A recent paper by Horton et al. (2021) using Italian data fails to find evidence that mandatory audit firm rotation improves audit quality when controlling for partner rotations. Using the Spanish setting (mandatory rotation from 1988-1995), Ruiz-Barbadillo et al. (2009) also find no significant impact of audit firm rotation. Findings from studies on perceptions of audit firm rotation provide are similarly mixed (e.g., Reid & Carcello, 2017; Horton et al., 2018). The results of studies investigating the potential effects of voluntary auditor changes on factual audit quality are also inconsistent (e.g., DeFond & Subramanyam, 1988; Fargher et al., 2008; Firth et al., 2012; Huang et al., 2015). Finally, a significant body of archival research has investigated the relationship between audit firm tenure and factual audit quality, again with mixed findings (e.g., Deis & Giroux, 1996; Myers et al., 2003; Knechel & Vanstraelen, 2007; Stanley & DeZoort, 2007; Singer & Zhang, 2018; Garcia-Blandon et al., 2019). However, some archival studies show lower audit quality in the initial years of engagement, consistent with a loss of auditor competence after auditor change. Furthermore, some studies reveal an inverse u-shaped relationship between auditor firm tenure and audit quality (Davis et al., 2009; González-Diaz et al., 2015; Brooks et al., 2017).⁴

Some experimental studies show that mandatory audit firm rotation improves auditor independence (e.g., Dopuch et al., 2001; Arel et al., 2006; Jennings et al., 2006; Gates et al., 2007; Hatfield et al., 2011; Aschauer & Quick, 2018). Other experiments fail to identify a significant effect (e.g., Kaplan & Mauldin, 2008; Kamath et al., 2018; Quick & Schmidt, 2018).

3.2.3

Joint audits

Joint audits could potentially reduce the economic bonding between the auditor and the client because audit fees are shared between two firms. In addition, two auditors may be more likely to resist client pressure, which should improve auditor independence (Zerni et al., 2012). Finally, the four-eyes-principle and the associated review of the work of the other audit firm potentially strengthens auditor independence and audit quality (Ittonen & Trønnes, 2015; Lesage et al., 2017). However, joint audits may cause a free-riding problem, i.e., one audit firm could shirk its responsibilities and save audit resource costs by reducing its own efforts and relying on the other audit firm's work (Deng et al., 2014; Holm & Thinggaard, 2016). Furthermore, joint audits could be associated with coordination problems and insufficient information exchange between the joint auditors. Therefore, the overall effect of joint audits on audit quality is again theoretically unclear.

Prior research on the impact of joint audits is primarily archival and thus measures the effect on factual and perceived audit quality without being able to isolate the relationship between joint audits and auditor independence. In most cases, discretionary accruals are used to proxy factual audit quality. The results are mixed. Evidence from France and Denmark on mandatory joint audits failed to identify a significant impact on factual audit quality (André et al., 2016; Lesage et al., 2017). By contrast, findings from Sweden and Finland with a voluntary joint audit setting showed a positive effect on factual and perceived audit quality (Zerni et al., 2010; Zerni et al., 2012; Itonnen & Trønnes, 2015). Similarly, Al-Hadi et al. (2016) report a negative association between joint audits and the cost of debt for the Gulf Cooperation Council countries.

Another stream of research investigated the optimal auditor-pair choice in a joint audit setting, again with mixed findings. While Francis et al. (2009) and Bennouri et al. (2015) find that the use of two Big 4 audit firms results in the highest audit quality, Lobo et al. (2017) document that a Big 4-Non-Big 4 auditor pair results in the most extensive goodwill impairment, when low-performance indicators suggest such an impairment.

Only a few studies directly analyzed the association between joint audits and factual or perceived auditor independence. Ruhnke & Schmidt (2016) surveyed German managers and supervisory board members, revealing a positive effect on their independence perceptions (but an overall negative effect). Conversely, an experimental study with German bank directors and institutional investors conducted by Quick & Schmidt (2018) indicated a negative impact on perceived auditor independence. Likewise, Hoos et al. (2019), who performed an experiment with Dutch auditors, found that joint audits resulted in lower professional skepticism in their judgment of the going concern assumption.

3.2.4

Critical evaluation

Archival research investigating the effects on factual audit quality must apply audit quality proxies (Le et al., 2022). Thus, the main weakness of this research strand is associated with such surrogates. Using qualified audit opinions is problematic because it assumes that all qualifications reflect higher audit quality. However, unmodified audit opinions only indicate poor audit quality if the financial statements are not free from material misstatements. If an unmodified audit opinion is appropriate, it will represent high audit quality. The application of going-concern opinions is limited to financially distressed firms; however, limiting the analyses to bankrupt companies would be more precise. Both proxies are associated with rare events, and the opinions could be false, which limits their applicability (Francis, 2004). The most common audit quality proxy is earnings management. However, earnings management is only estimated and serves as a surrogate of accounting quality, which proxies audit quality (Knechel et al., 2013). In addition, earnings management is only sometimes associated with an infringement of accounting standards and, thus, does not always indicate low audit quality. Moreover, accruals-based earnings management activities reverse over time, and it is typically unobservable for researchers to what extent discretionary accruals are driven by accruals reversals and what the time lag of those reversals is. Restatements are a direct output-related audit quality measure but are rare events limiting applicability.

Archival studies investigating capital market reactions are the dominating research method for analyzing the impacts on perceived audit quality. However, capital markets are permanently flooded with information (Habib, 2012). Thus, it is difficult to control for other disclosed information than audit and NAS fee, audit tenure, or joint audit information, as well as for other contemporaneous events occurring on or around filing dates (Kinney & Libby, 2002). Furthermore, the influence of audit and financial reporting quality on, e.g., firm value is relatively small compared to other factors. Consequently, there is a high risk of potentially omitted correlated variables (DeFond & Zhang, 2014).

A caveat of many experimental studies is using students as a proxy for non-institutional investors. Moreover, there is no guarantee that experiment participants behave the same way as in real-life situations. For example, the amount of information included in experimental materials must be much less than in reality, and consequently, the variables of interest could be overstressed. Additionally, most experimental studies lack evidence of construct validity (Asay et al., 2023). Survey respondents can easily identify the research objective, which may result in biased answers.

A final concern is that most prior research investigated the effects on audit quality. Thus, these studies only consider the net effect of, e.g., auditor-provided non-audit services, that is, the difference between a potentially positive effect on auditor competence and a negative effect on auditor independence. Related inferences on auditor independence, therefore, are only indirect.

3.3

Research questions

Despite the regulatory initiatives outlined above, auditor independence is still questioned (Carmichael, 2023) and is the focus of regulatory and oversight bodies. For example, the Public Company Accounting Oversight Board recently introduced a new section of the inspection reports focusing on independent violations (PCAOB, 2023).

Against the backdrop of the limited number of measures for strengthening auditor independence discussed by regulators and addressed by research and the fact that prior discussion has focused on PIE audits, this study uses an explorative approach to identify alternative measures suggested by non-professional investors and auditors to enhance auditor independence for PIE, but also non-PIE audits. Prior research mainly stems from Anglo-Saxon countries. Therefore, the study surveyed Spanish subjects and raises the following research questions:

RQ1: What independence-enhancing measures do Spanish non-professional investors suggest for PIE audits?

RQ2: What independence-enhancing measures do Spanish auditors suggest for PIE audits?

RQ3: What independence-enhancing measures do Spanish non-professional investors suggest for non-PIE audits?

RQ4: What independence-enhancing measures do Spanish auditors suggest for non-PIE audits?

4

Methodology

This study is based on a survey. For the development of the questionnaire, prior experimental research and surveys on auditor independence including innovative independence enhancing measures, discussions in front of recent related legislative reforms, statements from professional organizations, and publications from or on the auditing profession published in practitioner journals, all related to independence-enhancing measures, were analyzed. This was supplemented by brainstorming among the authors of this paper and members of their working groups. The preliminary questionnaire was then validated through 15 expert interviews with ten auditors, a lawyer, a financial analyst, and three audit committee members. The interviewees confirmed their suitability for most of the measures included in the preliminary questionnaire. However, some additional suggestions emerged during the interviews and resulted in an expansion of the questionnaire. Only a few measures were withdrawn from the questionnaire.

Subsequently, non-professional investors pilot-tested the questionnaire to ensure that it was clear, understandable, and consistent. They resulted in only minor adjustments and a reduced number of analyzed measures, because the pilots regarded the questionnaire as too long.

The final questionnaire encompasses 50 potentially independence-increasing measures. Six of them are only related to audits of non-PIE, because they are already implemented for PIE audits. The participants were asked to assess the impact of these measures on auditor independence on a 7-point Likert scale (from 1 = measure is entirely unsuitable to 7 = measure is perfectly suitable), separately for PIE and non-PIE audits. For some measures which are related to a time restriction (e.g., regarding rotations) or a monetary restriction (e.g., fee caps), the participants were asked to make a quantitative assessment (in years or as a percentage).

Regarding the selection of participants, the intention was to reflect the suppliers and the users of auditing services. Auditors participated in the survey because of their high expertise in auditing and accounting. Non-professional investors are strongly affected by information asymmetries and mainly must rely on audited financial statements, whereas professional investors have access to additional sources of information, like analyst conferences. Thus, there is a much more vital need to protect the former. Moreover, their perceptions are crucial to capital markets’ efficiency and effectiveness (Kaplan & Mauldin, 2008). The official Spanish register of auditors kept by ICAC was accessed to identify auditor names, explored the Internet for their e-mail addresses, and invited 500 auditors to participate in the survey. This resulted in 62 usable responses from auditors and a response rate of 12.4 percent.

In addition, the Prolific platform was used to select 250 non-professional investors from Spain. Prolific enables targeting a representative sample of non-professional investors who participated on a first-come-first-serve basis. Compared to samples used in many prior studies, Prolific's participant pool is older and has a range of education levels and employment experience. Furthermore, there is growing evidence that findings using Prolific participants are comparable to studies acquiring participants in classic ways (Crump et al., 2013; Palan and Schitter, 2018). This resulted in 252 usable responses from non-professional investors.⁵

For reliability analysis, Cronbach's Alpha was calculated to assess the internal consistency of the questionnaire. The results (auditor responses for PIE audits = 0.952; auditor responses for non-PIE audits = 0.834; investor responses for PIE audits = 0.950; investor responses for non-PIE audits = 0.955) suggest a high or even excellent internal consistency.

Based on the assumption that late respondents are roughly equivalent to the perceptions of non-respondents, t-tests for potential non-response bias were applied by comparing early with late responses (Oppenheim 2000) for both participant groups, auditors, and investors. The means differ significantly in only a few cases (5 out of 188), indicating that such a bias is not an issue and supporting representativeness of the study. The survey was administered online using SoSci Survey.

The participating auditors have, on average, nearly ten years of auditing experience, and the proportion of male and female auditors is almost equal. Sample's private investors have owed shares for about four and a half years and have an average level of auditing knowledge. They are predominantly male. For both subject groups, the majority are younger than 30. Table 1 provides details on the demographics.

Table 1.

Demographics

AuditorsFor how many years have you been worked in auditing services?N = 62Mean = 9,77SD = 10,885Min. = 1Max. = 46GenderN = 62Female = 28Male = 34AgeN = 62< 30 = 3030 – 39 = 1140 – 49 = 1250 – 60 = 9

Private-InvestorsGenderN: = 252Female = 77Male = 174Diverse = 1AgeN = 252< 30 = 13630 – 39 = 5540 – 49 = 4850 – 60 = 0> 60 = 13For how many years do you owe shares?N = 252Mean = 4.52SD = 6.903Min. = 0Max. = 47Knowledge of auditing (7 point Likert scale, 1 = low; 7 = high)N = 251Mean = 3.40SD = 1.415Min. = 1Max. = 7

The mean and the standard deviation for each item were calculated and, due to their common use, t-tests were applied to investigate whether the means differ significantly from the scale's midpoint. Measures with a significantly higher mean than four were classified as suitable for improving auditor independence. The applied scale is ordinal, which means that typical prerequisites like the normality of data distribution and equality of variance in standard deviation are violated. Nonetheless, t-tests are robust to ordinal distributions for larger sample sizes (Boneau, 1960).⁶ Furthermore, t-tests were applied to investigate whether the perceived need for independence-enhancing measures differs between PIE and non-PIE, and between non-professional investors and auditors. Finally, regressions were run to analyze whether participants’ demographic characteristics are significantly associated with their evaluation of independence-enhancing measures:⁷

MEASURES_inv = β₀ + β₁ GENDER + β₂ AGE + β₃ YEARS-SHARE + β₄ KNOW-AUDIT + ε

MEASURES_aud = β₀ + β₁ GENDER + β₂ AGE + β₃ YEARS-AUDIT + ε

5

Results

5.1

Non-professional investors’ perceptions of independence-enhancing measures

Table 3 summarizes the survey results regarding non-professional investors. The majority of highly ranked measures are related to oversight, controls, and sanctions. This can be explained by the theories of moral hazard and risk compensation. Individuals will take higher risks, and the likelihood of moral hazard increases, if they assume that third parties will suffer the potential consequences of such risks (theory of moral hazard). Furthermore, people adjust their behavior in response to perceived levels of risk, becoming more careful where they sense greater risk and less careful when feeling more protected (theory of risk compensation; e.g., Lavym & Miller, 2000). Thus, auditors will have incentives to reduce their independence level and increase their risk exposure if they do not bear the full costs of that risk, e.g., if they are not unlimitedly liable for damages resulting from non-independent behavior. Therefore, improved oversight, sophisticated controls, and strict sanctions may incentivize higher audit quality and foster auditor independence.

The deterrence theory (Becker, 1968) is based on the notion of rational decisions (Schell-Busey et al., 2016). A decision-maker violates standards if the expected benefit exceeds the resulting costs (Cornish & Clarke, 1987). Thus, auditors choose the alternative for which the most significant perceived gap exists between perceived benefit and perceived costs. Accordingly, deterrence from the threat of punishment is multifaceted, that is, it depends on the certainty, severity, and swiftness of punishments. This explains investor preferences not only for strict sanctions but also for effective control systems.

Legitimacy theory may explain why investors regard control system external verifications as highly valuable. Legitimacy theory explains that companies assume an implicit social contract between themselves and society, which, in turn, encourages them to comply with certain norms and values (Deegan, 2019). Accordingly, audit firms react to community expectations by implementing legitimating strategies, such as whistleblowing and compliance management systems (Patten, 2020). Obtaining third-party assurance can be a valuable tool for addressing concerns regarding the credibility of such implemented systems. Hence, such assurance may enhance the credibility of the verified systems and thereby legitimize the audit firm's related efforts (Simnett et al., 2009).

Mandatory whistleblowing systems (rank 1 for PIE, rank 1 for non-PIE) for audit firms and their external verification (rank 3, rank 4) are the measures mostly appreciated by investors. Whistleblowers are individuals who detect and report misconduct in an organization. Whistleblowing systems offer channels for individuals to report information they observed about potential wrongdoing and provide them with protection from retaliation. Research and observations from practice have shown that whistleblowing can effectively contribute to faster detection and resolution of misconduct (e.g., Dyck et al., 2010). The mere fact that whistleblowing increases the likelihood of detecting misconduct contributes to prevention. Thus, investors perceive such systems, which are a mechanism for avoiding and reducing the possibility of independence violation, as an effective approach for enhancing auditor independence. Similar arguments apply to mandatory compliance management systems for audit firms (rank 16, rank 19).

Compliance management systems encompass a set of processes and measures to protect enterprises from possible regulatory compliance violations (Gammisch & Balina, 2014). Compliance management is essential for audit firms because perceptions of their integrity, ethics, and independence are fundamentally important. Audit firm managers have incentives to signal their organizational commitment to ethical principles and the law to the public (Ferell et al., 2017). Investments in compliance management systems can keep audit firms proactive regarding continuous improvement processes, and flexible regarding regulatory changes (Perskow, 2003; Anon et al., 2007). The high ranking of whistleblowing systems may also be caused by the EU Whistleblowing Directive from 2019, which increased public awareness of this topic (European Union, 2019). Likewise, the proposal for an amended EU Sustainability Reporting Directive (European Union, 2021) stresses the relevance of non-financial information, e.g., on corporate governance, which whistleblowing systems serve, and the need for related assurance services. This could explain the appreciation of external verifications of audit firms’ whistleblowing and compliance management systems (rank 11, rank 11). An entity's whistleblowing and compliance management systems are mainly unobservable to the public. Consequently, the public bears an information risk regarding the existence and effectiveness of such systems. Thus, the importance of assurance in this context as a specific type of investment becomes evident (Jennings, 2012). ISAE 3000 (revised) on assurance engagements other than audits or reviews of historical financial information provides a framework for such services.

The EU Regulation on statutory audits of PIE (European Union, 2014b) already requires a competent authority for the surveillance of the activities of PIE auditors (Article 20). This authority is responsible for quality assurance at least every third (PIE audits) or every sixth (non-PIE audits) year (Article 26). Such inspections assess the design of statutory auditors' internal quality control system. For testing compliance, audit files are selected, based on an analysis of the risk of failure to carry out audits adequately. Against this backdrop, it is surprising at first glance that investors perceive improvements in public oversight as a highly effective measure for fostering auditor independence (rank 2, rank 2). However, the current system still leaves room for improvement. For example, the control cycle could be shortened, and the sample size of reviewed audit files could be increased. Moreover, there is evidence that inspectors are not well paid. The ICAC inspectors are public servants with a basic annual salary of between 24,365 and 31,514 Euros. It is questionable whether it is possible to attract appropriately qualified inspectors with such salaries. Thus, better remuneration could also increase this function's attractiveness and improve the quality of inspections. Furthermore, the ICAC employed just 18 inspectors in 2021, so that better and more personnel resources would also enhance public oversight. A final possible improvement would be the disclosure of the inspection reports, which would enable the users of audited financial statements to evaluate both the audit quality provided by an auditor and auditor independence.

Additionally, investors view specific external independence controls (rank 7, rank 8), in addition to external quality controls and internal independence controls (rank 10, rank 12), as effective tools for ensuring auditor independence. According to the International Standard on Quality Control 1, the audit firm must establish and maintain a system of quality control that includes policies and procedures that address, among other things, the relevant ethical and, more specifically, independence requirements (ISQC 1.16), and ensure that the firm and its staff maintain independence, e.g., the communication of independence requirements to staff, the identification, evaluation, and handling of independence threats (ISQC 1.21.). This requires an ongoing and mutual flow of information between the audit firm, engagement partners, and staff (ISQC 1.22). Similarly, the engagement partner shall make a decision on compliance with independence requirements that apply to the audit engagement (ISA 220.11). Article 29 of the EU Audit Directive requires the Member States to ensure that all statutory auditors and audit firms are subject to an external quality assurance system, and Spain has implemented this. These quality assurance reviews focus on audit firms' internal quality control systems, but their scope also includes an assessment of compliance with independence requirements (European Union, 2006; European Union, 2014a). The same holds for the inspections according to Article 26 of the EU Regulation on statutory audits of PIE (European Union, 2014b). In summary, internal and external audit quality controls already exist and include independence controls. However, non-professional investors believe that separate internal and external independence control systems are more effective.

Another measure that non-professional investors expect to be very effective is an increase in civil liability (rank 4, rank 7). In 2008, the EC recommended limiting the civil liability of statutory auditors against the audited company and third parties, arising from a breach of their professional duties, except in cases of intentional violation. It mentioned three methods for limiting liability, a liability cap, the application of proportional instead of joint and several liability, and a contractual limitation of liability (European Commission, 2008; for a discussion, see Doralt et al., 2008; Samsonova-Taddei & Humphrey, 2015). Further suggestions for limiting liability, namely a fixed or variable cap related to client size or audit fees, were not considered by the European Commission (London Economics & Ewert, 2006). Spain opted for proportionate liability (Article 26, Law 22/2015). The methods for limiting liability offer possibilities for tightening auditor liability exposure, e.g., by applying joint and several liability instead of proportionate liability. In addition, a reversal of the burden of proof, which means that the auditor must prove that liability conditions (misconduct, fault, damage, and causality) are not fulfilled, would also exacerbate civil liability. Finally, courts could be less restrictive regarding fulfilling the prerequisites for third-party liability. The rationale behind a stricter liability regime is that auditors are exposed to a higher liability risk or higher liability claims, leading to more conservative behavior and consequently higher independence of mind. Prior archival research has already revealed a positive relationship between auditor liability exposure and audit quality (Francis & Krishnan, 2002; Gaver et al., 2012; Firth et al., 2012; He et al., 2017).

Similarly, non-professional investors regard the introduction of specific criminal sanctions for auditors as a meaningful measure for improving auditor independence (rank 5, rank 3). Some jurisdictions have special legislative provisions that threaten statutory auditors with imprisonment in cases of specific misconduct. In Germany, e.g., statutory auditors can be imprisoned up to three in cases of intentional violation of auditors’ professional secrecy or five years in cases of intentionally forming an incorrect opinion on whether the financial statements are prepared in accordance with the applicable financial reporting framework (paragraphs 333 and 332 HGB). Spanish investors suggest similar provisions in cases of intentional violation of independence standard. The deterrence effect of such provisions would be powerful because imprisonment would have severe financial consequences and result in exclusion from the profession.

Another type of harsher sanction that is highly appreciated by investors is related to disciplinary sanctions (rank 6, rank 13), which potentially protect the public interest (Bédard, 2001). Disciplinary sanctions have a repressive and a preventive function (Blij et al., 1998). For the latter, it is not sufficient that a balanced set of sanctions be available, e.g., reprimands, fines, suspensions, and exclusion from the profession/the register of statutory auditors. It is also necessary that such a range of sanctions is applied in a balanced way. It is a strength of the Spanish system that the oversight body publishes the cases, including the names of sanctioned auditors or audit firms because this increases the preventive effect of sanctions (Quick & Warming-Rasmussen, 2002). Table 2 provides an overview of the published sanctions imposed by the ICAC in Spain from 2017 to 2022. It shows that only a minority of infractions were classified as very serious and that the misconducts were sanctioned with a fine in most cases. Only sixteen cases related to the violation of independence standards during this period. Thirteen of them were sanctioned with a fine, one of about 10.5 million Euros and another one of about 2 million Euros. The other fines were marginal, i.e., less than 50,000 Euro. Only two cases resulted in a suspension of two respectively three years. In summary, the Spanish oversight body tends to impose mild sanctions and use more severe sanctions only in a few extreme cases. Such a pattern reduces the preventive effect of disciplinary sanctions.⁸ Thus, deterrence could be strengthened if the ICAC imposed more severe sanctions. Moreover, the publications from Spain do not reveal details on the type of misconduct, such as the specific independence standard which was violated. Therefore, clients and users of audited financial statements are not fully informed about the independent behavior of auditors.

Table 2.

Published Sanctions in Spain 2017-2021

Year	Nr. of sanctions	Very serious	Serious	Minor	Suspension	Fine
2017	75	1	65	9	5	70
2018	65	6	54	5	8	57
2019	68	0	68	0	3	65
2020	27	0	23	4	1	26
2021	43	0	40	3	3	40
2022	62	4	58	0	2	60
Total	340	11	308	21	22	318

Table 3.

Nonprofessional Investor Independence Perceptions

A final measure related to sanctions is the contractual arrangement on a penalty payment in case of non-independence between the audit firm and client, including disclosing the amount of penalty payment (rank 9, rank 10).

Another measure is assessed as highly appropriate for enhancing auditor independence in both cases, audits of PIE and audits of non-PIE, namely independence ratings for audit firms conducted by an independent rating agency (rank 12, rank 14). This would reduce the risks to the providers of capital. Auditors would be incentivized to act independently because a negative ranking could negatively affect their reputation and result in lower audit fees, a loss of clients, or difficulties in acquiring new clients. Concerning PIE audits, investors perceive the public tendering of audit engagements as a highly effective instrument (rank 8). The tendering process is regulated in Article 16 of the EU regulation but is only mandatory for PIE audits in conjunction with auditor changes. The primary argument for audit tendering is a potential increase in competitiveness within the audit market. Above and beyond that, however, it could reduce the length of the auditor-client relationship and thereby lower familiarity, avoid bonds of loyalty or emotive relationships, and contribute to auditor independence (Allam et al., 2017). However, the balance of power could shift in favor of the client, and accordingly, public tendering may also weaken auditor independence (Boon et al., 2005).

Non-professional investors perceive three of the measures already implemented for PIE audits as highly beneficial for non-PIE audits. The publication of non-audit service fees (rank 6) helps users to assess the risk of financial dependence and provides rough insight into the type of auditor-provided services. An independence declaration of the auditor to the audit committee or other people at the client company charged with governance before accepting or continuing an audit engagement (rank 9) is likewise valued. In addition, non-professional investors clearly prefer transparency reports (rank 5). According to Article 13 of the EU Regulation, a statutory auditor or an audit firm that carries out statutory audits of PIE must publish an annual transparency report, at the latest four months after the end of each financial year, which includes a statement concerning the statutory auditor's or the audit firm's independence practices, which also confirms that an internal review of independence compliance has been conducted. Another component of a transparency report is the total revenue earned by the statutory auditor or the audit firm, divided into revenue from PIE audits, from non-PIE audits, from permitted NAS to audit clients, and from non-audit clients (Pott et al., 2008; Deumes et al., 2012; La Rosa et al., 2019).

Only a few measures were not accepted by the surveyed non-professional investors, namely measures related to an internal rotation of audit partners, participation of the auditor in the general assembly, and the prohibition of tax consulting services provided by the auditor. Regarding the latter, the opinions of investors revealed by this survey correspond with prior research findings, which showed that the impact of tax consulting service fees on independence of mind and in appearance differs from the effect of other NAS. Often, a positive relationship was found (e.g., Christensen et al., 2015; Eilifsen et al., 2018; Watrin et al., 2019; Castillo-Merino et al., 2020). Regarding non-PIE audits, the prohibition of auditor-provided other assurance services and further prohibitions of other consulting services are also not accepted.

5.2

Auditor perceptions of independence enhancing measures

Table 4 provides information on auditor perceptions of the suggested independence-improving measures. Surprisingly, the number of instruments suggested by auditors is relatively high (39 out of 44 for PIE audits; 37 out of 50 for non-PIE audits). While non-professional investors prefer indirect measures related to oversight, controls, and sanctions, auditor perceptions are more diverse, and they also rank some direct instruments relatively high. In addition, compared to non-professional investors, auditor suggestions for PIE audits differ more from those for non-PIE audits.

Table 4.

Auditor Independence Perceptions

Similar to non-professional investors, auditors assess some measures related to oversight and controls as highly beneficial: further improvement of public oversight (rank 1 for PIE, rank 1 for non-PIE), internal independence controls (rank 3, rank 2), mandatory whistleblowing systems for audit firms (rank 5, rank 5), external independence controls (rank 6, rank 20), and compulsory compliance management systems for audit firms (rank 8, rank 4). However, auditors have lower preferences for sanctions, probably due the direct and indirect economic consequences of such sanctions. Nevertheless, they rank highly a contractual arrangement of a penalty payment in case of non-independence between the audit firm and client (rank 4, rank 6) and, surprisingly, stricter civil liability of auditors and audit firms (rank 9, rank 12).

Furthermore, auditors seem to recognize that not only audit firm and audit partner tenure, but also the tenure of audit team members, may cause a familiarity threat, resulting in an impairment of independence. Familiarity threats may occur due to close relationships with the client, client representatives, or client staff, deteriorating skepticism about potential misstatements, leading to an increased likelihood of a stale audit program, and a decreased chance that the auditor will make decisions contrary to the prior year's decisions, and the use of less rigorous audit procedures, or even to collusion with management (Carey & Simnett, 2006; Corbella et al., 2015; Patterson et al., 2019). Consequently, auditors suggest the introduction of mandatory rotation for audit team members after a certain period (rank 11, rank 17) and a cooling-off period after the rotation for audit team members (rank 2, rank 9) of about three years.

The provision of NAS is already limited in many jurisdictions, mainly due to self-review threats. The EU Regulation includes a longer black-list of prohibited NAS regarding PIE audits. Non-audit service fees increase the total revenue earned from a client and, thus, self-interest threats. Moreover, they may cause familiarity and advocacy threats (Hohenfels & Quick, 2020). For the users of audited financial statements, it would be helpful to be informed about auditor-provided NAS in order to assess auditor independence. Currently, many jurisdictions, including the EU, require the disclosure of auditor-provided non-audit service fees, differentiated into total fees for other assurance services, total fees for tax advisory services, and total fees for other NAS. However, prior research has shown that the impact of different NAS belonging to the category of other NAS on independence of mind and in appearance varies. This explains auditors’ appreciation of a disclosure of details on auditor-provided NAS (rank 10, rank 7). Fee dependence may also exist if an audit firm earns audit fees from one client, representing a high proportion of the audit firm's total revenue (e.g., Craswell et al., 2002). Therefore, auditors suggest a related limit (rank 7, rank 19).

A final instrument prioritized by auditors in conjunction with PIE is joint audits (rank 12, rank 23), where two different audit firms are appointed, who share the audit work and jointly sign the audit report (European Commission, 2010). Reciprocal control and review of working papers may improve independence. Furthermore, collusion would be more difficult for clients because they would then have to collude with two audit firms (Ratzinger-Sakel et al., 2013).

Regarding non-PIE audits, auditors also prefer two other measures, internal independence training (rank 8) and the introduction of a statutory fee schedule for audit fees (rank 11). The latter could contribute to a more independent and transparent fee-setting process. Downside fee pressure could increase incentives to sell NAS to audit clients, which, in turn, may threaten auditor independence. Unusually high audit fees could also threaten auditor independence, because clients could be using them in effect to pay for dependent auditor behavior (Choi et al., 2010). Furthermore, abnormally high audit fees increase the economic bond between auditor and client, posing another independence threat.

The questionnaire also included some measures which are already implemented for PIE audits. Thus, the related questions had only to be answered for non-PIE audits. The auditors express a high preference for two of them, namely transparency reports (rank 3) and an obligation of the auditor to declare independence to the audit committee or others at the client who are charged with governance before accepting or continuing an audit engagement (rank 10).

Auditors did not accept five measures regarding PIE audits, most related to auditor-provided NAS, and 13 measures in conjunction with non-PIE audits. Moreover, they rejected two measures for non-PIE audits, more prohibited NAS and a complete prohibition of NAS.

5.3

Additional analyses

A series of additional analyses were performed (untabulated). A first set of tests checked for significant differences between PIE and non-PIE audits. The means of non-professional investors and auditors are, for all measures, higher for PIE audits than non-PIE audits. There is only one exception related to non-professional investors. For the measure ‘shorter period for internal rotation of the responsible audit partner/introduction of internal rotation’, the mean for PIE audits (4.14) is slightly lower than the mean for non-PIE audits 4.15), but the difference is insignificant. In most cases (34 out of 44), the non-professional investors’ mean responses are significantly higher for PIE audits (t-tested). The pattern is similar for the auditors’ responses, i.e., for 30 measures, the means are significantly higher for PIE audits. This indicates a more substantial need for enhancing auditor independence for PIE audits than for non-PIE audits.

Second, t-test tested for significant differences between non-professional investors and auditors’ responses. Regarding PIE audits, auditors perceive most measures, i.e., 28, as more important than non-professional investors. The latter rate only 15 measures as more important than auditors do. However, the differences in the means are, in most cases, insignificant. For only seven measures, the mean differences are significant, and for six of them, the mean for the group of auditors is significantly higher. This is surprising because auditors have to bear the related costs of the measures, whereas investors most likely perceive them as free commodities. This finding may be explained by the greater expertise of auditors concerning their independence. The pattern for non-PIE audits is different. Non-professional investors have, on average, a slightly higher preference for the suggested independence-enhancing measures than auditors. Again, the number of measures with significant mean differences between the two groups is low. Out of six measures with significant mean differences, four are valued more by non-professional investors.

Third, regressions were run to investigate whether there is a significant relationship between the participants’ demographic characteristics and responses. The results reveal that female and younger investors and auditors perceive a significantly lower need for improving auditor independence. Investors with longer shareholder experience and less knowledge of auditing show significantly stronger support for independence-enhancing measures. The more experienced auditors are, the less their demand to foster independence.⁹

6

Discussion

Independence is a cornerstone of the audit profession, and auditors are permanently suspected of not acting independently. In the aftermath of financial crises and accounting scandals, regulators have tightened measures to strengthen auditor independence, mainly focusing on the prohibition of NAS and the rotation of audit firms. Nevertheless, the public still frequently assumes that integrity, objectivity, or professional skepticism has been compromised.

Against this bedrock, this study aimed to identify alternative regulatory measures that could improve independence of mind and appearance. This aligns with prior research findings on the impact of NAS fees on auditor independence (Quick, et al., 2023), which demonstrate that non-professional investors are particularly skeptical regarding auditor independence. The study's findings show that non-professional investors perceive a greater need for further independence-enhancing measures than auditors. Non-professional investors typically base their decisions solely on information from audited financial statements, so these addressees must be particularly protected. Therefore, regulators must pay particular attention to their perceptions. Another more general finding of this research is that non-professional investors and auditors presume a greater need to improve auditor independence in the case of PIE audits. This result is unsurprising because agency conflicts are more profound in PIEs than non-PIEs. However, the survey participants also suggest measures to protect auditor independence for non-PIE audits. In contrast, recent regulatory actions focused on PIE audits.

Regarding concrete measures, the study reveals that those related to oversight, control, and sanctions are perceived as most promising by both groups, non-professional investors and auditors. Thus, implementing and external verification of whistleblowing and compliance management systems and improving public oversight could be promising measures. A more severe civil liability, stricter disciplinary sanctions, and the introduction of specific criminal sanctions for auditor misconduct or penalty payments in case of non-independence are especially painful for auditors. Nevertheless, not only non-professional investors but also auditors have a strong preference for such measures. However, auditors' preference for sanctions is lower than that of non-professional investors. Presumably, both parties assume a direct and positive relationship between the severity of such sanctions and the independent behavior of auditors. Auditors also suggest a mandatory audit team rotation, which would probably be fine because such rotation often happens automatically. Moreover, auditors would appreciate joint audits. Auditors from non-Big 4 audit firms were likely the drivers of this finding because these audit firms would benefit from mandatory joint audits.

The study's results are of potential interest to regulators when searching for further independence-enhancing measures and for audit firms, which could voluntarily implement such measures as a signal to markets.

7

Conclusion

External audits serve the public by increasing the credibility of financial reports. However, this function can only be fulfilled if audit quality is sufficiently high, which requires auditor independence. Despite numerous regulatory initiatives to ensure auditor independence and audit quality, the number of accounting scandals has recently increased. In addition, regulatory change was essentially limited to the prohibition of NAS, the capping of non-audit service fees, and the introduction of mandatory audit firm and audit partner rotation. Likewise, prior research focused on NAS, audit firm rotation, the impact of auditor tenure on audit quality, and joint audits.

Against this backdrop, this survey-based explorative study aimed to identify alternative independence-enhancing measures perceived by Spanish non-professional investors and auditors. Using a list of 50 measures that may improve auditor independence, the results indicate that most of the suggested measures are perceived as beneficial by both non-professional investors and auditors. Non-professional investors particularly favor measures related to oversight, controls, and sanctions. Auditors also perceive measures about oversight and control as helpful but rank most sanctions lower than do non-professional investors. There is one exception. Surprisingly, auditors also prefer stricter civil liability. In contrast to non-professional investors, auditors rank not only indirect but also some direct independence-improving measures high, e.g., the introduction of mandatory rotation for audit team members and a related cooling-off period. Unexpectedly, auditors perceive a more substantial need to implement measures fostering auditor independence for PIE audits.

The findings are of interest to regulators, auditors and audit committees, and researchers. Recent accounting scandals have demonstrated that insufficient regulatory measures have been taken so far. Thus, regulators who perceive a need to implement additional measures for improving auditor independence may find promising suggestions. Such actions may reduce the likelihood of distortions in the capital market and increase the credibility of audited financial statements. Audit firms and clients could gain a competitive advantage by voluntarily implementing measures to signal auditor independence. Research findings offer auditors and audit committees the opportunity to identify effective measures for voluntary and proactive actions to ensure auditor independence and high stakeholder perceptions of audit quality. Audit firms could signal a reduced moral hazard risk and help clients avoid adverse selection. Thereby, auditors might be more able to attract new clients, retain current clients, and claim higher audit fees. Clients could apply such measures to signal auditor independence to their shareholders, which could trigger positive capital market reactions, e.g., a reduction of the costs of equity. Prior research has focused on prohibiting non-audit services, audit firm and audit partner rotation, and joint audits. Due to the explorative nature of the study, the results can help researchers identify promising avenues for future research on auditor independence.

This study is not free from limitations. First, the application of a survey enables participants to identify the research objectives, which may result in biased responses, e.g., investors may be tempted also to suggest measures of which they are not convinced, because they assume that they do not have to bear the costs of such measures. Second, the study only surveyed two groups, non-professional investors and auditors. Other users, like financial analysts, bankers, or audit committee members, may have different perceptions. Third, participants are from Spain only. The Spanish environmental setting is indeed similar to other Continental European countries but different from Anglo-Saxon countries. Finally, the study ignored the costs associated with implementing independence-enhancing measures, which may exceed the benefits of some of the measures investigated.

CRediT authorship contribution statement

Reiner Quick: Writing – review & editing, Writing – original draft, Project administration, Methodology, Formal analysis, Data curation, Conceptualization. Daniel Sánchez Toledano: Writing – review & editing, Writing – original draft, Project administration, Methodology, Formal analysis, Data curation, Conceptualization. Joaquín Sánchez Toledano: Writing – review & editing, Writing – original draft, Project administration, Methodology, Formal analysis, Data curation.