Multinational audit firms frequently present themselves as warriors of honesty, due care and ethical conduct by conveying messages that: (1) their hands are clean (i.e. they are not to blame for audit failure), (2) their hands were tied (i.e. they were powerless to prevent audit failure), (3) their work was good and (4) their intentions are good (Dunne, Brennan & Kirwan 2021:101170). Paradoxically, fraud and corruption scandals, excessive greed and weak governance structures persistently plague corporations and government institutions worldwide, even though regulatory audits are conducted (Shore 2021). Sadly, auditors themselves have also been embroiled in professional misconduct, collusion, dishonesty and even fraud, which have resulted in some being fined by regulators or expelled from the audit profession, and a few even being sent to prison (Cowan 2024; IRBA 2024a; McCance 2019; Ramsarghey & Hardman 2020).

Auditing remains the primary method for reducing information asymmetry and limiting fraud and corruption by strengthening governance and oversight (Alsughayer 2021; Campa, Quagli & Ramassa 2025; Dakhli 2021; Zahid et al. 2022). As a result, there is a direct nexus between audit quality, which enables accountability and transparency, and financial stability at auditees, as defined in auditing theory (Kausar, Shroff & White 2016; Nikmah & Arjoen 2023; Zahid et al. 2022). Consequently, improved audit quality results in more credible and reliable financial statements, enhanced trustworthiness and reduced self-serving behaviour.

Globally, governments regulate the auditing profession for several reasons, and preventing transgressions is undoubtedly one of them. Audit regulators use sanctions to, inter alia, deter auditors from improper conduct, in accordance with deterrence theory, in addition to using sanctions to punish auditors for reckless behaviour, ethical violations and non-compliance with international auditing standards (Dharmasiri et al. 2022). Deterrence theory presumes that auditors will always act rationally and will only act improperly or transgress if it is worthwhile and profitable to do so (Cheung & Jia 2024; De Puiseau, Glöckner & Towfigh 2019; Henson, Nguyen & Olaghere 2023). The probability of non-compliance with auditing standards and improper conduct escalates with the potential savings and benefits to the auditor and, theoretically, should decline with a high probability of detection and associated severe sanctions. In accordance with deterrence theory, the IRBA’s primary mandate is to conduct inspections at audit firms to verify the auditors’ compliance with the ISAs and IRBA Code and issue sanctions in instances where transgressions are detected (IRBA 2025a).

Complementary to deterrence theory, proportionality theory dictates that sanctions should be just and fair, and therefore there should be an appropriate equilibrium between the monetary value of the sanction and the gravity of external harm caused by the transgression (Dachak 2021; Gruodytė & Orlauskas 2024; Raskolnikov 2019). This theory is foundational to the criminal justice system and is applied by the judiciary when sentencing offenders. However, proportionality theory has been criticised for causing unnecessary significant harm to respondents and for resulting in disparities in sanctioning, leading to unequal punishment for similar transgressions (Brown Coverdale 2025). Moreover, Ríos Carrillo (2023) accentuated that proportionality theory faces problems of measurability and comparability, as it may not always be practical to measure and compare the trade-off between the monetary value of sanctions and the value of external harm caused by transgressions. In this regard, a large multinational audit firm was sanctioned R200 000 for failure ‘to comply with the requirements of the International Standard on Quality Control (ISQC) 1, as non-compliance with the ISAs was identified across several engagement files’ in June 2025 (IRBA 2025a:14). In such instances, it is challenging to apply proportionality theory, as the external harm caused by this transgression is incommensurable.

There is no doubt that the fear of severe punishment will deter rational auditors from improper conduct and non-compliance with the ISAs, according to deterrence theory. However, deterrence theory has also been criticised for focusing on the severity of punishment rather than on rehabilitation (Du Preez & Muthaphuli 2019), and for the fact that harsh penalties do not stop instances of improper conduct and non-compliance with the ISAs. In this regard, it was observed that although the IRBA significantly increased the monetary value of fines from 15 June 2023, the number of charges against auditors doubled from this date onwards, contradicting both deterrence and proportionality theory (Refer to Table 5 in the empirical section of this study). A possible explanation for this phenomenon is that misconduct and non-compliance with the ISAs often result from a complex business environment, complicated accounting and auditing standards, low audit fees, a stressful work environment, inadequate communication, urgent reporting deadlines and demanding audit clients, which may compromise the registered auditor’s ability to execute a fully ISA-compliant audit (Abrahams & Phesa 2025; Harber et al. 2023). These factors are not entirely within the auditor’s control and therefore cannot be eliminated by increasing the severity of fines for misconduct.

The external auditing ‘brain drain’ has profoundly troubled audit firms and regulators internationally (Ghasemi, Mehrabanpour & Talebnia 2023). Ma et al. (2024) found that there is a direct correlation between high auditor attrition rates and audit quality failures, which will likely result in audit clients having to restate their annual financial statements. Prior period errors and restatements are deemed by investors and regulators as the most reliable indicators of audit quality failures. High auditor attrition rates also result in audit firms losing skilled and experienced auditors, which has a detrimental impact on audit quality (Hu et al. 2025).

In the United States of America, the number of accountants and auditors declined by 17% from 2019 to 2022, with more than 300 000 professionals leaving the profession. Reasons for this decline have been attributed to burnout, low salaries and difficulty in passing the Chartered Public Accountant’s examinations (Ellis 2022). Similarly, the Australian Securities and Investments Commission also reported a 16.8% decline in the number of Registered Company Auditors from 2019 to 2024 (Australian Securities and Investments Commission 2025). In the United Kingdom, the number of statutory audit firms declined by 24.9%. Likewise, a decline was also observed in the number of individuals holding the audit qualification over the past decade because of an ‘image problem’ as high-profile auditing failures such as Carillion and Wirecard have tarnished the public’s trust in auditors, in addition to auditors experiencing a highly stressful work environment and lower salaries in comparison with finance professionals (Prosser 2025). The external auditor ‘brain drain’ is a global phenomenon and is mostly driven by burnout, complex regulatory burdens and an exodus of young talent to other sectors. This talent flight threatens to compromise audit quality, intensify workloads for remaining staff and increase audit fees globally.

Audit regulators demand flawless audits. Consequently, external auditors use quality management mechanisms such as engagement quality reviews, partner peer reviews, monitoring reviews, manuals, checklists, technical consultations and the automation of auditing processes to prevent and detect audit errors and to avoid enforcement sanctions by regulators (Gold et al. 2022). The auditing profession must minimise audit errors to maintain public credibility and legitimacy (Beck, Nicoletti & Stauber 2022). Nevertheless, despite the profession’s best efforts to prevent audit errors, they occasionally occur because auditors are human and prone to making mistakes.

The Auditing Profession Act, No.26 of 2005 (APA), empowers the IRBA to implement any necessary measures to protect the public in their interactions with registered auditors (RSA 2005). These steps include initiating action to promote the integrity of the auditing profession through investigating allegations of misconduct by registered auditors, conducting disciplinary hearings and imposing sanctions (RSA 2005). It is hoped that enforcement action will discourage registered auditors from committing similar transgressions henceforward (IRBA 2025c).

According to the IRBA’s Fine Guideline, audit transgressions that occurred prior to 15 June 2023 could attract a maximum fine of R200 000 per charge. As of 15 June 2023, fines were increased. Individual registered auditors are subject to fines ranging from a minimum of R100 000 to a maximum of R10 million per charge, while audit firms are subject to fines ranging from a minimum of R200 000 to a maximum of R25m per charge (IRBA 2025c). On 06 June 2024, the IRBA indicated that the threshold for minimum fines would no longer apply, while the maximum fines remained unchanged (IRBA 2025c). The consequences and penalties for transgressions can be significant and severe; therefore, it is imperative that auditors ethically comply with the ISAs to avoid the IRBA’s disciplinary processes and sanctions.

Stemming from deterrence and proportionality theory, it is expected that IRBA sanctions would have declined after 15 June 2023, as higher fines are expected to discourage auditors from committing transgressions. Thus, the nature of charges the IRBA raised against respondents, such as breaches of the IRBA Code, the inability to obtain sufficient appropriate audit evidence and insufficient monitoring processes, is likely to influence the value of fines imposed. A detailed analysis of the nature of these transgressions, as well as the types of sanctions imposed by the IRBA in response, will reveal the hidden relationship between the impact of increased fines on the subsequent transgressions raised by the IRBA. The following hypothesis is formulated:

The empirical sections of this study examine the impact of transgressions on fines and cost orders imposed by the IRBA. Initially, a retrospective document analysis of 405 disciplinary cases against registered auditors was conducted. The study included all cases where sanctions were imposed by the IRBA, as reported in quarterly IRBA newsletters from 01 January 2019 to 30 June 2025. The IRBA newsletters were obtained from the IRBA website for the period under review. Every sanction entry included the following: (1) the name of the respondent, unless the IRBA ruled that the identity of the respondent should be withheld, (2) IRBA’s description of the charges against the auditor or audit firm, (3) monetary value of the fine imposed, (4) value of the cost order imposed and (5) descriptions of required addition remedial actions. The dataset used in this study represents the known population of IRBA sanctions imposed from 01 January 2019 to 30 June 2025.

A dataset was compiled in ATLAS.ti version 25, ATLAS.ti, Berlin, Germany, from the charges as published in the quarterly IRBA newsletters. Slightly more than one-third (325/923, or 35.2%) of charges referenced a specific transgression relating to a specific auditing standard, accounting standard or section of a law for which a registered auditor received a sanction. If the IRBA explicitly references a transgression, no further hand coding was performed for that charge. For sanctions for which no specific law or transgression was explicitly stated, the researcher coded the transgressions based on the narrative description of the charges drafted by the IRBA. Furthermore, the IRBA often raised more than one charge against a registered auditor, yet only one sanction was imposed against an auditor or audit firm for all charges combined. As such, the number of charges exceeds the number of sanctions that auditors had received. In instances where more than one charge was raised against an auditor, the associated charges were coded separately, even if only a single combined sanction was issued for all the separate charges. Transgressions associated with non-compliance with ISAs, International Standard on Quality Management (ISQM), ISQC standards and legislation may result in IRBA Code violations simultaneously. This results in a research design limitation because of dependence and intersection between variables, which may have an associated impact on the research results.

A single coder performed all the coding to ensure the coding instrument was reliable and consistent, rather than using multiple coders. The accuracy, completeness and validity of the coding were confirmed by three independent, experienced registered auditors from separate audit firms who specialise in conducting engagement quality reviews. The reliability of this study was enhanced by ensuring that the selected content categories pertained to specific ISAs, International Financial Reporting Standards (IFRS), IRBA Code, laws and regulations. For measuring the agreement among coders, Krippendorff’s alpha is a reliability coefficient, drawing distinctions among typically unstructured phenomena. As a result, Krippendorff’s alpha = 1, indicating perfect agreement among raters, and confirms the reliability of the coding. Cohen’s kappa and Gwet’s AC1 could not be used to measure agreement among coders, as they evaluate the agreement between binary ratings. Table 1 presents a summary of IRBA charges in this dataset before and after hand coding.

Figure 1 indicates that the most frequently sanctioned transgressions were breaches of the IRBA Code and instances of non-compliance with ISA 500, which related to the auditors’ inability to obtain sufficient and appropriate audit evidence. The rest of the transgressions were more evenly spread. More about the prominent transgressions and spread will be dealt with further in this study.

The sanctions imposed by the IRBA were classified into the following categories: (1) Value of the fine; and (2) value of the cost order, which were clearly described in the IRBA newsletters, and professional judgement was not required to manually code the nature of sanctions imposed by the IRBA. The combined total of fines and cost orders was expressed in Rands and adjusted for the impact of time value of money devaluations. Consumer Price Inflation for South Africa as reported by the World Bank Group (2026) was used to adjust the historical values of fines and cost orders from 2019 to 2024 to be comparable, by using 2025 as a base year. The following time value of money formula was used to perform this calculation: FV = PV × (1 + R), where FV represents the future value, PV represents the present value and R represents the discount rate, which in this case represents the annual Consumer Price Inflation, and was analysed as the dependent variable in SPSS version 30, IBM, Armonk, New York, United States. Refer to Table 2 for a summary of fines and cost orders for each sanction.

As per Figure 1, all transgressions with the ISAs, IRBA Code and legislation, as stipulated in the IRBA’s charges, were classified as independent variables (refer to Appendix 2 Table 1-A2 for a list of the independent variables). Each observation included in this study, therefore, represents a separate charge raised against an auditor. These charges were then aggregated based on the nature of each transgression with a particular ISA standard, section of the IRBA Code and legislation as specified in the IRBA News. Some auditors were found guilty of multiple transgressions at the same time; in these instances, each charge was counted separately. In the analysis, independent variables were expressed as whole numbers based on the cumulative number of transgressions for each variable as reported in the IRBA News from January 2019 to June 2025. In total, 75 independent variables were observed. More information on the definition and contents of the ISAs, IRBA Code and relevant legislation can be found at: https://www.irba.co.za/guidance-to-ras/technical-guidance-for-ras.

Multicollinearity: Multicollinearity assesses the degree of interrelationship among the independent variables included in a regression model. To evaluate potential multicollinearity among the transgression and non-compliance variables, eigenvalues derived from the collinearity diagnostics were examined. Small eigenvalues approaching zero may indicate that one or more dimensions of the predictor matrix contain limited unique information and may therefore signal multicollinearity. The results showed that 84% (63 of the 75) dimensions exhibited relatively moderate to large eigenvalues, suggesting that severe multicollinearity was not pervasive among the independent variables. These findings indicate that the various categories of transgressions and non-compliance generally contributed distinct information to the model. It should be noted that eigenvalue diagnostics assess relationships among the independent variables only and do not provide evidence regarding the relationship between the transgression variables and the value of fines imposed by the IRBA.

From a standardised regression analysis perspective, data points (representing various sanctions) form a distinct pattern and are tightly clustered around the centre, suggesting relatively small and consistent residuals for most transgressions, representing a normal distribution. Certain data points were far from the central cluster, indicating that one outlier resulted in the model predicting a much lower fine and cost order than the actual value. Two have extremely high predictive values and small residuals. However, the data point pattern suggests that the regression residuals are mostly homoscedastic, but the model is influenced by a few extreme outliers that may distort regression estimates. A brief overview of the outlier cases is provided below:

The association between pertinent charges and the monetary value of fines and cost orders was examined using multiple linear regression analyses. Hypothesis testing included multiple linear regression, R² (coefficient of determination) analysis. The multiple linear regression model used in this study is as follows (Equation 1):

where

Ethical clearance to conduct this study was obtained from the College of Business and Economics Research Ethics Committee of the University of Johannesburg (No. SAREC20250821/07).

The primary aim of this study is to identify common audit transgressions to empower auditors to avoid the mistakes of the past. As indicated in Figure 1, registered auditors most often received sanctions for breaches of the IRBA Code (170 charges, or 18.4%). Code violations most often relate to the IRBA’s concern that auditors are no longer independent when conducting audits. It was observed that charges related to instances where the auditors failed to implement adequate safeguards to reduce threats to their independence to an appropriately low level, as such, instances in which auditors prepared and audited the same set of annual financial statements for their clients, as well as instances in which the same audit firm provided its clients with non-assurance services in addition to conducting regulatory audits, attracted sanctions. Therefore, auditors must ensure compliance with independence requirements, as threats to independence may also compromise professional scepticism.

The second most common cause for sanctions is the auditors’ inability to obtain sufficient and appropriate audit evidence, as stipulated in ISA 500 (154 charges, or 16.7%). Frequently, auditors omitted to provide evidence that their auditees’ revenue recognition and measurement practices comply with the prescriptions of IFRS 15 and that performance obligations have been met before revenue is recognised. In these instances, the IRBA would concurrently assess the impact of any revenue misstatements on total materiality to determine whether the audit opinion remains appropriate. Moreover, transgressions were linked to concerns about the sufficiency of audit evidence for group consolidations and the elimination of intercompany revenue transactions, as required by IFRS 3.

The third most common source of IRBA sanctions is an audit firm’s failure to comply with the International Standards on Quality Control (ISQC 1, which was revised and re-issued as ISQM 1), resulting in 60 charges (6.5%). Often, the IRBA will sanction an audit firm when audit quality across several engagements was insufficient, raising doubts about the leadership team’s ability to obtain reasonable assurance that the audit firm has complied with the ISAs and consistently conducts high-quality audits. The leadership team of the audit firm serves as the custodians of audit quality across all audit engagements conducted by the audit firm. Repetitive IRBA inspection findings regarding a lack of independence in several audit engagements will prompt the IRBA to simultaneously assess whether the audit firm has complied with ISQC 1 (which was subsequently revised and reissued as ISQM 1).

While it is not surprising that most auditors received sanctions for breaches of the IRBA Code, ISA 500 and ISQC 1, non-compliance with IAS 7 resulted in more specific charges (36 charges or 3.9%). The IRBA expected auditors to identify errors regarding the inclusion of non-cash items in their clients’ cash flow statements, which constituted a breach of an accounting standard rather than non-compliance with a particular auditing standard. The IRBA dictates that auditors should be familiar with relevant accounting standards such as IFRS, IFRS for Small and Medium Enterprises (SMEs), Generally Recognised Accounting Practices (GRAP) and Modified Cashflow Standards (MCS) in addition to the ISAs, and ensure that their clients comply with accounting standards in all material aspects; alternatively, the auditors should issue a modified audit opinion.

An analysis of the annual rate of sanctions imposed by the IRBA from 01 January 2019 to 30 June 2025 is reflected in Table 3.

As presented in Table 3, the annual number of sanctions more than doubled from 40 in 2019 to 84 in 2024. In contrast, the number of sanctions in 2020 may have been lower, stemming from the consequences of lockdowns during the COVID-19 pandemic. The data, however, suggest a growing trend in the number of sanctions imposed by the IRBA annually, despite the dwindling number of registered auditors.

Using the total number of registered auditors from the IRBA annual reports for each year, Table 3 presents a relative measure of disciplinary actions per thousand registered auditors. Interestingly, the number of sanctions per thousand registered auditors has more than doubled from 2019 to 2024, despite the IRBA’s audit engagement inspections declining from 136 in 2019 to 109 in 2024. It can be inferred that the recent higher sanctions rates from 2021 to 2025 may reflect increased transgressions, changes in enforcement, regulatory changes (including the introduction of ISQM) or other factors, as the number of IRBA audit engagement inspections conducted declined during the same period. This phenomenon seems to contradict deterrence theory as the increase in sanctions did not result in a reduction in transgressions.

The IRBA issued fines for 98.0% of sanctions (397/405). In total, the IRBA issued 397 fines, with the average fine costing registered auditors R211 839.00 during the period under review. Additionally, 6.4% (26/405) of respondents were issued with cost orders, with an average price tag of R665 853 per order, as depicted in Table 2. In March 2024, the IRBA issued its largest fine to an individual respondent, amounting to R2 150 000 for audit failures and improper conduct related to the Steinhoff International audit. The charges included references to the lack of safeguards to reduce threats to independence to an acceptable level. Furthermore, the respondent was accused of failing to perform the appropriate audit procedures, as required by the ISAs, to identify and assess the risks of material misstatement for the group and to obtain sufficient and appropriate audit evidence from component auditors. The respondent was also accused of insufficient involvement in the work of component auditors, among other allegations.

The most significant cost order was imposed in March 2023, when the respondent was ordered to pay IRBA R4 932 119.00 for legal fees incurred, after expressing an incorrect audit opinion for state-owned entity Eskom. The respondent also failed to disclose Eskom’s material non-compliance with legislation and significant internal control deficiencies in Eskom’s audit report. Apart from the cost order, the respondent was fined R200 000 and required to participate in mandatory public sector auditing training. The total costs the respondent would have incurred for this transgression are significant, comprising IRBA’s cost order for legal fees, the respondent’s own legal fees, the fine due and payable to IRBA and training costs.

Table 4 indicates that while the IRBA usually withheld respondents’ names, it named them in 132 published cases. The IRBA may exercise its discretion in terms of Section 51(5) of the APA to intentionally publish the names of respondents who are guilty of improper conduct or to report on the transgression in general terms. Of the 132 respondents whose names were published in IRBA newsletters, two respondents resigned from the IRBA before the conclusion of the IRBA investigations process, and the names of an additional 44 registered auditors no longer appear in the IRBA registry. Of the 132 sanctioned individuals with publicly known identities, an estimated 59 (44.7%) are no longer registered auditors as of 01 September 2025. Furthermore, 13 registered auditors have been permanently prohibited from registering with the IRBA and from practising as auditors. Independent Regulatory Board for Auditor’s sanctions for individual registered auditors often result in the Registered Auditor no longer practising as an auditor after receiving a sanction, thereby increasing auditor attrition rates and potentially leading to a further reduction in the total number of registered auditors.

Panel A of Table 5 presents the absolute counts of auditing, accounting or legal transgressions, along with the type of sanction imposed, for each year from 2019 to 2025. An analysis was subsequently performed to compare the data across the periods preceding and following the IRBA fine guideline change. Comparisons of the period prior to the IRBA fine guideline change indicate a lower average fine of R214 557.00 compared to the period after the IRBA fine guideline change, when the average fine increased to R264 422.00, representing a 23.2% increase. Contrary, the value of cost orders decreased from R815 234.00 for the period prior to the fine guideline changes to R706 885.00 afterwards, representing a 13.3% decline, which was caused mainly by significant cost orders being raised by the IRBA for the charges against auditors relating to the VBS Bank and Eskom audits and Nkonki Incorporated, in the period prior to the IRBA fine guideline change.

Panel B of Table 5 compares annual charges for ISAs, IFRS and legal transgressions from 2019 to 2025, before and after the IRBA fine guideline change. Notably, this analysis revealed that overall transgressions increased by 68.3% from the period preceding the IRBA fine guideline change to the period thereafter. Specifically, charges relating to breaches of the IRBA Code increased by 66.4%, whereas non-compliance with ISA 500 increased by 92.1%, and breaches of ISQC 1 (which was subsequently revised and reissued as ISQM 1) requirements increased by an astounding 141.1%. This research study discovered compelling evidence that the average number of charges increased in correlation with higher fines following the implementation of the new IRBA fine guideline document on 15 June 2023, which strongly contradicts deterrence theory. This phenomenon may be because of auditors being more rigorously regulated by the IRBA.

In view of the considerable number of independent variables, the adjusted R² is the figure that should be used, as the number of variables inflates the unadjusted R². The adjusted R² corrects for the number of predictors and the sample size. The regression model (adjusted R² = 0.345) is moderately strong in social and regulatory research contexts (Ozili 2023), indicating that the type of the transgression is a meaningful factor in determining fines and cost orders, although other unmeasured factors (such as severity within each type, number of offences, mitigating circumstances, past record, enforcement discretion or randomness) account for the remaining 65.5%. The type of transgression is a meaningful determinant of fine and cost order size. The model’s explanatory power is moderate, as it captures a moderate portion of the factors that influence fines and cost orders, but not all of them.

Fine and cost order sizes are reasonably (but not perfectly) aligned with the types of transgressions. A lower adjusted R² does not, in itself, imply inconsistency in sanctioning outcomes, as the unexplained variance may be attributable, inter alia, to omitted variables, differences in case-specific circumstances or limitations in the model specification. As indicated in Figure 2, the adjusted R² of 0.345 indicates some structure but still room for variability.

The ANOVA section tests whether the model significantly predicts fine size. Since p < 0.01, it can be concluded that the set of transgression types collectively provides a statistically significant prediction of fine or cost order size, as indicated in Figure 3. This would confirm that at least one transgression type significantly contributes to explaining the variation in fine or cost order sizes.

Coefficients (Individual Predictors): Each variable (e.g. ISA 500, ISA 705, IRBA Code, Section 45 of the APA, etc.) has a coefficient (β) and significance value (Sig.). These tell you which types of transgressions are statistically significant predictors of higher or lower fines and cost orders.

Interpretation pattern:

From the correlation section visible in the output (each r value in the correlation table represents the simple bivariate relationship between one specific transgression and the total fine or cost order amount, without controlling for the others), several types exhibit noticeable positive correlations with total fines or cost orders, such as:

These variables are likely to emerge as key predictors in the coefficients table (Table 6). The nature of professional misconduct plays a significant role in determining the amount of fines or cost orders. The analysis suggests that fine size is primarily determined by the types of transgressions committed. Half of the differences in fine amounts can be attributed to whether an auditor violated specific standards (such as the ISAs or IRBA code). International Standards on Auditing 200 and ISA 220 also show moderate positive correlations. However, substantial unexplained variation remains, indicating that qualitative or contextual factors (such as case severity, cooperation and repeat offences) also play a significant role.

Ten individual transgressions (13% of independent variables) (e.g. ISA 705, ISA 200 and ISA 560) reveal a weak to a very weak positive correlation with fine or cost order size, but when all types of transgressions are combined, the model explains a moderate variation in fines (adjusted R² = 0.345). Thus, the type of transgression matters, as some violations are more strongly associated with larger fines. However, no single transgression fully explains the size of a fine or cost order. It is the combination of multiple types of transgressions that provides a moderate (34.5%) explanation of how fines are determined.

Based on the results of several statistical tests, this study signifies that transgressions such as breaches of the IRBA Code and APA and non-compliance with the requirements of ISA 705, ISA 200 and ISA 315 are somewhat positively correlated with the monetary value of fines and cost orders. Therefore, in testing the hypothesis, it can be established that transgressions associated with non-compliance with the ISAs, IFRS and legislation do not all equally influence the monetary value of fines and cost orders similarly and in the same proportion. The moderately strong linear regression model has therefore rejected the hypothesis. The fact that most transgressions relating to non-compliance with the IRBA code displayed a very weak positive correlation to the value of fines, and cost orders may be a possible explanation for why increases in the IRBA’s enforcement efforts and fines did not result in an associated decline in charges as expected, as predicted by deterrence and proportionality theory. Albeit other factors, such as enforcement intensity, case mix, reporting lags or other confounding factors, may also have contributed to a very weak positive correlation between transgressions and the value of fines and cost orders.

Since only about a third of IRBA transgressions cited specific standards or regulations, it was necessary to hand-code the rest. Furthermore, the IRBA newsletters did not specify which financial year-end audit each charge referred to. Therefore, it is plausible that auditors only received sanctions in 2025 for audits conducted many years before. In this regard, the complete and timely reporting on the details of sanctions is important and will enable effective monitoring and transparency.

Untabulated findings indicate that IRBA sanctions are commonly associated with grievances raised by audit clients, members of the public and employees within audit firms (IRBA 2024a). In this respect, it is advisable to assess the incentives and disincentives associated with IRBA complaints. An evaluation of the impact of IRBA sanctions on audit firms’ sustainability may also be an interesting research topic. Similarly, an international comparison between transgressions in South Africa versus other jurisdictions will also provide valuable insight.