ESG Performance, Auditor Choice, and Audit Opinion: Evidence from an Emerging Market

Abstract

This study examines the effect of environmental, social, and governance (ESG) performance on auditor choice and audit opinion for Egyptian-listed firms. We use univariate and multivariate analyses of 612 firm-year observations for a sample of 68 firms listed on EGX100 over 2014–2022 using binary logistic regression models. Consistent with the ethical perspective of corporate social responsibility, we found that firms listed in the ESG index are more likely to assign one of the Big4 auditors, and less likely to receive a qualified opinion. Through an additional analysis, we found that COVID-19 moderates the relationship between ESG performance, auditor choice, and audit opinion. Our results confirm the value of ESG performance for audit practices in emerging economies. This research indicates that ESG performance can enhance financial reporting quality. Further, it ensures that binding guidelines and regulations are crucial to oversee corporate ESG performance, especially during crisis times, and enhance investors’ protection and firms’ sustainability.

1. Introduction

There is a recent growing awareness concerning the value of environmental, social, and governance (ESG) reporting and performance, and that ignoring ESG disclosure could negatively affect a firm’s image and, hence, its market value [1,2]. Recent firms’ scandals and collapses worldwide have motivated modern enterprises to consider ESG seriously when formulating or reviewing their strategies [3,4]. The majority of the literature has examined the financial implications of ESG performance and disclosure (e.g., [5,6,7,8,9,10,11]). However, it is also crucial to examine the strategic, not only financial, effects of ESG [12,13,14]. Along with this direction, recent studies have examined the implications of ESG performance and disclosure for the information environment and transparency [15]. That is, with ESG, it is anticipated that firms would act ethically considering the interests of all stakeholders [16,17], which can enhance their information environment and augment their transparency level [18,19]. Against this background, ESG might have implications for audit-related decisions, practices, and outcomes [20]. We believe that firms’ ESG performance and disclosure could have consequences for their auditor choice and the outcome of the audit process. Our study objective is twofold. It examines the extent to which the firms listed in the ESG index act ethically concerning financial reporting, by assigning higher-quality or Big-4 auditors. Further, it investigates whether these firms provide more accurate accounting information, and hence, they are less likely to receive a qualified audit opinion. In doing so, we focus on a developing market—the Egyptian stock market.

As opposed to Western developed markets, developing ones are more subject to the impact of severe information asymmetry and agency costs [21,22,23]. This situation is due to several reasons, including the existing weak legal systems [24] and ineffective corporate governance mechanisms in these markets [25,26]. In such developing contexts with higher information asymmetry, auditors tend to play a vital role [27,28,29]. An external audit can act as a mechanism to protect investors’ interests by uncovering managerial manipulation by adjusting both income-increasing and income-decreasing accruals [30,31,32,33].

We address the Egyptian market because, in recent years, there has been an increasing focus on the importance of ESG practices in its business sector. The government of Egypt has recently implemented policies and regulations to promote sustainable development and encourage firms to adopt responsible business practices (see [34,35]). For example, the Ministry of Investment and the Egyptian Institute of Directors introduced the corporate governance code in 2005 as guidelines for firms listed on the Egyptian Stock Exchange. In 2016, the Egyptian Financial Supervisory Authority replaced the existing governance code with revised detailed governance rules to be applied by both listed and unlisted firms. The new rules highlighted the value of disclosing nonfinancial information, including ESG, to consider the interests of all stakeholders [36]. Another important landmark in enhancing sustainability performance in the Egyptian market was the application of the S&P/EGX ESG index in 2010. This index identifies the best-performing companies listed in EGX 100 concerning environmental, social, and corporate governance activities and reporting [5]. In 2019, the Egyptian Ministry of Manpower and Migration and the Egyptian Ministry of Environmental Affairs issued Law No. 12 to promote sustainable development by protecting workers’ rights and enhancing work conditions. In the same year, the government formulated its national strategy to stress the importance of having a sustainable and responsible business environment. Further, in its 2020 report, the Egyptian Ministry of Investment and International Cooperation stressed the importance of increasing the share of renewable energy in the country’s energy mix. In 2021, the Egyptian Financial Supervisory Authority issued its guidelines for environmental and social risk management for the business sector. Such guidelines aimed to reduce the environmental and social risks associated with operational and financial activities [37]. However, the real impact of these activities is minimal due to the lack of a binding legal system that monitors and enforces compliance with the existing sustainability-related rules and regulations [36]. This unique context has motivated us to examine the implications of such a new context for audit-related practices—namely, auditor choices and the audit process outcome.

This study contributes to the literature concerned with the implications of ESG performance on auditor choice (AC) and audit opinion (AO). To our knowledge, this is the first study that examines the implications of ESG for AC and AO in Egypt as an influential emerging economy in the Middle East and North Africa region. Binary logistic regression (BLR) analysis showed that ESG firms are more likely to assign a Big4 auditor, and these firms are less likely to receive a qualified audit opinion. Moreover, as additional analyses, we considered the effects of COVID-19 on the relationship between ESG performance, AC, and AO. The results indicated that COVID-19 moderates the relationships between ESG performance, AC, and AO. The findings revealed an increased probability of ESG firms assigning one of the Big4 auditors during COVID-19. However, the results suggested an increased likelihood of ESG firms receiving a qualified audit opinion during COVID-19. Our results provide significant evidence to policymakers responsible for formulating guidelines and regulations to oversee firms, enhance governance, and protect stakeholders’ interests, especially during crises. Further, in line with the current findings, investors are advised to consider ESG performance while making investment decisions, especially during crises such as COVID-19.

The remainder of the paper is structured as follows. Section 2 reviews the literature and develops the study hypotheses. Section 3 outlines the research methods. Section 4 presents and discusses the study findings. Finally, Section 5 concludes the paper.

2. Literature Review and Development of Hypotheses

The literature reveals two main streams regarding the perceptions and implications of ESG performance. On the one hand, it is anticipated that firms adhering to ESG performance and reporting have motives to be honest and act ethically concerning all stakeholders [17]. This is because ESG responsibility requires firms to meet the economic, legal, ethical, and voluntary expectations of society’s constituents [38]. In other words, commitment to ESG would require firms to treat stakeholders ethically or responsibly along with the principles of modern civil societies. This is consistent with the stakeholder theory postulating that corporate management should give equal attention to all stakeholders, rather than serving the interests of a particular group, namely shareholders [39]. ESG performance and disclosure is one way to achieve this social objective [40]. According to Branco and Rodrigues, social responsibility and reporting involve compliance with a set of ethical standards that govern the decision-making process within firms, in a way that limits harm to society or stakeholders [41]. Focusing on Turkey, Aslan and Şendoğdu found that social responsibility influences corporate ethical values and behaviors positively [42].

On the other hand, from an opportunistic perspective, ESG responsibility and reporting might be perceived by some companies as a means of greenwashing—that is, to polish their image [43] or to hide the negative or irresponsible behaviors of corporate management [44,45]. In this case, the ‘apparent’ ethical behavior of socially responsible firms would be mainly used as a tool by corporate management to attain some personal benefits; rather than benefiting all stakeholders [29]. In this regard, Prior et al. showed that social responsibility disclosure may be used as a means of managing legitimacy, by influencing public perception without a real positive change in the behavior of the entity [46]. For instance, it can be used to immunize corporate management that manipulates profits [46]. Hurst indicates that the presence of an ethical code and social policies in the firms does not necessarily guarantee the ethical treatment of stakeholders [47]. Lanis and Richardson showed that higher levels of social responsibility disclosures are associated with aggressive tax practices, which contribute to tax evasion [29]. Nirino et al. did not find a positive moderating influence of ESG concerning the association between controversies and financial performance [48].

2.1. Corporate Social Responsibility (CSR) and Auditor Choice

From an opportunistic perspective, companies may engage in ESG to offset their corporate irresponsible behaviors, enhance their corporate image, or gain legitimacy to operate [45,49]. It is believed that companies with such opportunistic perspectives will not commit to corporate ethical conduct and hence, they may not demand a higher audit quality. In this regard, using U.S. data, Lamptey et al. suggested that CSR activities may be associated with more audit complexities and risks [50].

In contrast, from the ethical perspective, it is believed that ESG performance and reporting would contribute to better accounting information quality [51]. This ethical perspective might induce companies to protect stakeholders’ interests by supporting audit quality [52,53]. In this regard, focusing on Indonesia, Handayati et al. found that firms audited by Big4 auditors are positively related to CSR disclosure [54]. Focusing on the French context, Dakhli found that the positive implications of corporate social responsibility are more obvious in firms audited by Big4 auditors [55]. Using the U.S. data, Sun et al. found that firms with higher CSR performance are more likely to engage industry specialist auditors [56]. Using international evidence, Hichri found a positive association between audit quality and integrated reporting [57]. Using U.S. data, Du et al. found that companies with higher CSR performance are likely to engage big (high-quality) auditors [52]. Following this perspective, we believe that firms with higher ESG performance and reporting are likely to support higher audit quality by engaging Big4 auditors. Thus, we set the first hypothesis as follows:

H1. 

ESG firms are more likely to assign big4 auditors.

2.2. ESG Performance and Audit Opinion

As previously indicated, according to the opportunistic perspective, firms may use ESG practices to hide some of the negative activities, including reporting irregularities [1,58]. From this stance, it is not anticipated to find a direct association between ESG performance and receiving an unqualified audit opinion. In this regard, Nguyen and Trinh found a non-linear influence of CSR on the likelihood of receiving unqualified opinions [59]. However, several studies in the literature support the ethical perspective of ESG, suggesting its positive implications for the outcome of the audit process.

According to the ethical perspective, firms committed to ESG practices are likely to be honest and trustworthy by having a strict code of ethics [57] that, in turn, would restrict reporting irregularities [60], and instead support transparency of financial reports and information quality [51,61]. Supporting this view, some studies revealed that firms’ ESG practices are related to lower misstatement and client business risk [60,62]. Similarly, other studies indicated that ESG firms are likely to be associated with fewer auditors and analysts forecasting errors [63], less auditor engagement risk [52], and less litigation risk [9].

This context makes us infer that firms committed to ESG may receive an unqualified audit opinion. Along with this argument, using evidence from energy firms listed in Vietnam, Nguyen and Trinh indicated that companies with noticeable CSR activity are anticipated to get unqualified opinions due to the quality of their financial reports [59]. Wang et al. indicated that firms’ engagement with ESG practices decreases the probability of receiving a qualified audit opinion [64]. Thus, we set the second hypothesis as follows:

H2. 

ESG firms are less likely to receive qualified audit reports.

3. Research Design

3.1. Sample and Data Sources

Our sample includes all the firms indexed in EGX100 across the period 2014–2022. We obtained the firms’ auditors and audit opinion data as well as financial data over the study period from the firms’ annual reports. Governance data were collected from the companies’ governance reports published by Egypt for Information Dissemination (EGID) Company. Finally, ESG performance data were collected through the ESG index published by the Egyptian Stock Exchange across the study period.

Table 1. The study sample.

	No. of Firms	Observations
Panel A: Sample selection
EGX100 across the study period 2014–2022	100 firms	900
(–) Banking and financial firms	(22)	(198)
(–) Firms with missing financial and governance data	(10)	(90)
Final sample	68	612
Panel B: Sample distribution
Merchandising	18	2.9%
Manufacturing	495	80.9%
Service	99	16.2%
Total	612	100%

shows the process of sample selection and sample distribution according to industry.

3.2. Research Models

Since the dependent variables (AC and AO) are dichotomous, we analyzed data depending on binary logistic regression (BLR). The maximum likelihood approach is used to estimate the model parameters, and an iterative algorithm is used to pick the coefficients that result in the most “likely” observation outcomes. Following Kurniawati et al. and Tantawy and Moussa [31,32], we used the following logistic regressions to test our hypotheses:

$$\begin{gathered} {AC}_{it}=\alpha +B_1\left({ESG}_{it}\right)+B_2\left({FSIZE}_{it}\right)+B_3\left({LEVERAGE}_{it}\right)+B_4\left({PROFITABILITY}_{it}\right)+B_5\left({FGROWTH}_{it}\right) \\ +B_6\left({LOSS}_{it}\right)+B_7\left({FAGE}_{it}\right)+B_8\left({BSIZE}_{it}\right)+B_9\left({BMEETINGS}_{it}\right)+B_{10}\left({DUALITY}_{it}\right) \\ +B_{11}\left({BINDEPENDENCE}_{it}\right)+B_{12}\left({ACSIZE}_{it}\right)+B_{13}\left({ACMEETINGS}_{it}\right) \\ +B_{14}\left({ACINDEPENDENCE}_{it}\right)+B_{15}\left(INDUSTRIES\right)+B_{16}\left(YEARS\right) \end{gathered}$$

(1)

$$\begin{gathered} {AO}_{it}=\alpha +B_1\left({ESG}_{it}\right)+B_2\left({FSIZE}_{it}\right)+B_3\left({LEVERAGE}_{it}\right)+B_4\left({PROFITABILITY}_{it}\right)+B_5\left({FGROWTH}_{it}\right) \\ +B_6\left({LOSS}_{it}\right)+B_7\left({FAGE}_{it}\right)+B_8\left({BSIZE}_{it}\right)+B_9\left({BMEETINGS}_{it}\right)+B_{10}\left({DUALITY}_{it}\right) \\ +B_{11}\left({BINDEPENDENCE}_{it}\right)+B_{12}\left({ACSIZE}_{it}\right)+B_{13}\left({ACMEETINGS}_{it}\right) \\ +B_{14}\left({ACINDEPENDENCE}_{it}\right)+B_{15}\left(INDUSTRIES\right)+B_{16}\left(YEARS\right) \end{gathered}$$

(2)

where AC_it is a binary variable showing whether the firm is audited by one of the Big-4 or non-Big-4 auditors. AO_it is a binary variable showing whether the auditor has issued a qualified or unqualified opinion. ESG_it is a binary variable showing whether firms are listed in the EGX index or not.

We follow some literature in controlling for some variables that influence independent variables, such as firm size, leverage, firm profitability, firm growth, loss, firm age, board size, board meetings, duality, nonexecutive directors in the board, audit committee size, meetings and independence, industries, and years [31,32]. All variables used in models (1) and (2) are reported in

Table 2. Variable measurements.

Variables		Measurement
AC	Auditor choice	A dummy variable assigned 1 for firms audited by a Big-4 audit firm in the year t, and 0 otherwise.
AO	Audit opinion	A dummy variable assigned 1 for firms that received a qualified opinion in the year t, and 0 otherwise.
ESG	ESG performance	A dummy variable assigned 1 for firms listed in the ESG index in the year t, and 0 otherwise.
FSIZE	Firm size	The natural logarithm of total assets in year t.
LEVERAGE	Financial leverage	Total debt over total assets in year t.
PROFITABILITY	Firm profitability	Net profit after tax, and extraordinary items in year t, scaled to total assets.
FGROWTH	Firm growth	The change in net sales in year t, scaled to revenue in year t − 1.
LOSS	Carryforward loss	A dummy variable assigned 1 if the firms have carryforward loss in year t, and 0 otherwise.
FAGE	Firm age	The natural logarithm of the number of years since the firm has been listed in the Egyptian Exchange.
BSIZE	Board size	The number of directors on the board in year t.
BMEETINGS	Board meetings	The number of board meetings in year t.
DUALITY	Duality	A dummy variable assigned 1 if the Chairman and CEO are the same person, and 0 otherwise.
BINDEPENDENCE	Board independence	The number of non-executive directors on the board, scaled to its total number of directors in year t.
ACSIZE	Audit committee size	The number of members in the audit committee in year t.
ACMEETINGS	Audit committee meetings	The number of audit committee meetings in year t.
ACINDEPENDENCE	Audit committee independence	The number of non-executive directors in the audit committee, scaled to its total number of directors in year t.

.

4. Research Results and Discussion

4.1. Descriptive Statistics

Panel A of

Table 3. Descriptive results.

Panel A: Descriptive results for all samples (Full Sample = 612)
Variables	Mean	Median		SD	Minimum	Maximum
AC	0.327	0.000		0.469	0.000	1.000
AO	0.413	0.000		0.493	0.000	1.000
ESG	0.240	0.000		0.428	0.000	1.000
FSIZE	21.050	21.138		1.728	16.822	25.639
LEVERAGE	0.432	0.433		0.248	0.001	0.986
PROFITABILITY	0.046	0.036		0.086	−0.124	0.225
FGROWTH	0.133	0.080		0.452	−0.848	1.057
LOSS	0.240	0.000		0.426	0.000	1.000
FAGE	2.910	3.044		0.546	0.693	4.080
BSIZE	8.173	8.000		2.797	3.000	16.000
BMEETINGS	10.005	9.000		4.628	2.000	23.000
DUALITY	0.716	1.000		0.451	0.000	1.000
BINDEPENDENCE	0.701	0.750		0.201	0.200	1.000
ACSIZE	3.583	3.000		0.987	2.000	8.000
ACMEETINGS	5.291	4.000		2.678	1.000	14.000
ACINDEPENDENCE	0.934	1.000		0.170	0.000	1.000
Panel B: Univariate analysis (Full Sample = 612)
Variables	Firms listed in the ESG index (147 observations)		Firms not listed in the ESG index (465 observations)		t-test	Sig.
	Mean	SD	Mean	SD
AC	0.646	0.480	0.226	0.419	10.239	0.000 ***
AO	0.211	0.409	0.477	0.500	−5.870	0.000 ***
FSIZE	22.255	1.515	20.669	1.614	10.534	0.000 ***
LEVERAGE	0.401	0.250	0.441	0.247	−1.711	0.088 *
PROFITABILITY	0.071	0.076	0.039	0.087	4.003	0.000 ***
FGROWTH	0.172	0.418	0.121	0.462	1.176	0.240
LOSS	0.068	0.253	0.294	0.455	−5.749	0.000 ***
FAGE	2.715	0.504	2.971	0.545	−5.049	0.000 ***
BSIZE	9.673	2.740	7.699	2.647	7.818	0.000 ***
BMEETINGS	10.347	5.140	9.897	4.455	1.028	0.304
DUALITY	0.612	0.489	0.748	0.434	−3.211	0.001 ***
BINDEPENDENCE	0.731	0.157	0.692	0.212	2.058	0.040 **
ACSIZE	3.707	1.218	3.544	0.899	1.753	0.080 *
ACMEETINGS	5.354	2.415	5.271	2.758	0.326	0.744
ACINDEPENDENCE	0.976	0.073	0.921	0.189	3.457	0.001 ***

* is significant at level ˂ 10%, ** is significant at level ˂ 5%, *** is significant at level ˂ 1%. For variable definitions, see Table 2.

shows descriptive results for the whole sample. The mean values of AC and AO are 0.327 and 0.413, indicating that 32.7% of firms assign one of the Big4 auditors, and 41.3% of firms receive a qualified audit opinion, which is consistent with previous studies such as Tantawy and Moussa [32]. The mean value of ESG is 0.240, indicating that 24% of our sample companies are indexed in the ESG index. Panel B shows the mean differences and t-tests for firms listed in the ESG index versus non-listed firms. The findings indicate that the ESG firms are more likely to assign one of the Big4 auditors and are less likely to receive a qualified audit opinion. These results are significant at the 1% significance level.

Table 4. Correlation matrix.

Variables	1	2	3	4	5	6	7	8	9	10	11	12	13	14	15	16
1-AC	1.000
2-AO	−0.380 ***	1.000
3-ESG	0.383 ***	−0.231 ***	1.000
4-FSIZE	0.337 ***	−0.082 **	0.392 ***	1.000
5-LEVERAGE	−0.055	0.175 ***	−0.069 *	0.036	1.000
6-PROFITABILITY	0.072 *	−0.093 **	0.160 ***	0.272 ***	−0.305 ***	1.000
7-FGROWTH	0.006	−0.048	0.048	0.090 **	0.042	0.135 ***	1.000
8-LOSS	−0.163 ***	0.229 ***	−0.227 ***	−0.349 ***	0.319 ***	−0.478 ***	0.037	1.000
9-FAGE	−0.256 ***	0.179 ***	−0.200 ***	−0.007	0.271 ***	−0.103 **	0.022	0.147 ***	1.000
10-BSIZE	0.140 ***	−0.118 ***	0.302 ***	0.414 ***	−0.116 ***	0.249 ***	0.014	−0.093 **	−0.074	1.000
11-BMEETINGS	−0.093 **	0.246 ***	0.042	0.167 ***	0.126 ***	0.191 ***	0.028	−0.089 **	−0.039	0.138 ***	1.000
12-DUALITY	−0.240 ***	0.191 ***	−0.129 ***	−0.093 **	−0.028	0.016	−0.006	0.074 *	0.004	0.017	0.202 ***	1.000
13-BINDEPENDENCE	0.197 ***	−0.289 ***	0.083 **	0.060	−0.211 ***	0.045	−0.021	0.014	−0.154 ***	0.379 ***	−0.096 **	−0.269 ***	1.000
14-ACSIZE	−0.055	0.180 ***	0.071 *	0.158 ***	−0.035	0.171 ***	0.028	−0.025	0.111 ***	0.273 ***	0.267 ***	0.164 ***	0.013	1.000
15-ACMEETINGS	−0.097 **	0.313 ***	0.013	0.013	0.067 *	−0.008	0.039	0.114 ***	0.142 ***	0.049	0.404 ***	0.108 ***	−0.233 ***	0.401 ***	1.000
16-ACINDEPENDENCE	0.197 ***	−0.187 ***	0.139 ***	0.035	−0.065	0.062	−0.016	0.005	0.006	0.175 ***	−0.099 **	−0.227 ***	0.289 ***	−0.184 ***	−0.186 ***	1.000

* is significant at level ˂ 10%, ** is significant at level ˂ 5%, *** is significant at level ˂ 1%.

presents a Pearson’s correlation matrix for our variables. ESG is positively (negatively) correlated with AC (AO) at the 1% level, respectively. This result supports the hypothesis that ESG firms focus more on financial reporting quality. Thus, ESG firms are more likely to assign one of the Big4 auditors and are less likely to receive a qualified audit opinion. These findings are consistent with some previous studies such as Gonçalves et al. and Wang et al. [64,65]. Further, the correlation matrix shows that independent variables are correlated below 0.5, which indicates that the multi-collinearity problem is non-existent.

4.2. Logistic Regression Results

We estimate models (1) and (2) depending on the binary logistic regression (BLR) method. The BLR models have met the requirements of the goodness of fit to predict AC and AO based on ESG performance, as presented in

Table 5. Testing the research hypotheses.

Panel A: Omnibus Test of Model Coefficients
Model (1): Auditor Choice					Model (2): Audit Opinion
Step 1	Chi-square	D.f.	Sig.		Chi-square		D.f.	Sig.
Step	247.868	24	0.000		205.309		24	0.000
Block	247.868	24	0.000		205.309		24	0.000
Model	247.868	24	0.000		205.309		24	0.000
Panel B: Hosmer–Lemeshow’s goodness-of-fit test
Model (1): Auditor Choice					Model (2): Audit opinion
	Chi-square	D.f.	Sig.		Chi-square		D.f.	Sig.
Step 1	9.455	8	0.305		11.310		8	0.185
Panel C: Nagelkerke R2 square test
Model (1): Auditor Choice					Model (2): Audit opinion
	−2 Log likelihood	Cox and Snell R2	Nagelkerke R2		−2 Log likelihood		Cox and Snell R2	Nagelkerke R2
Step 1	525.563	0.333	0.464		624.651		0.285	0.384
Panel D: Wald test
Variables	Model (1): Auditor Choice				Model (2): Audit opinion
	B (Wald)	Exp (B)	95% C.I. for Exp (B)		B (Wald)	Exp (B)		95% C.I. for Exp (B)
			Lower	Upper				Lower	Upper
Step1 ESG	1.127 *** (16.604)	3.088	1.795	5.311	−1.083 *** (14.193)	0.339		0.193	0.595
FSIZE	0.604 *** (43.296)	1.830	1.529	2.191	0.005 (0.003)	1.005		0.860	1.174
LEVERAGE	0.279 (0.233)	1.322	0.425	4.112	0.512 (0.952)	1.669		0.596	4.673
PROFITABILITY	−0.111 (0.005)	0.895	0.038	21.242	0.643 (0.194)	1.903		0.108	33.360
FGROWTH	0.050 (0.036)	1.051	0.632	1.748	−0.497 ** (4.330)	0.608		0.381	0.972
LOSS	−0.153 (0.200)	0.858	0.438	1.680	1.374 *** (21.576)	3.952		2.213	7.057
FAGE	−1.005 *** (19.871)	0.366	0.235	0.569	0.436 ** (4.494)	1.547		1.033	2.314
BSIZE	−0.122 ** (6.316)	0.885	0.805	0.974	−0.036 (0.594)	0.964		0.879	1.058
BMEETINGS	−0.092 *** (10.972)	0.912	0.864	0.963	0.094 *** (14.524)	1.098		1.047	1.152
DUALITY	−0.530 ** (4.273)	0.588	0.356	0.973	0.229 (0.859)	1.257		0.775	2.040
BINDEPENDENCE	2.712 *** (12.124)	15.057	3.272	69.294	−2.587 *** (15.724)	0.075		0.021	0.270
ACSIZE	0.047 (0.120)	1.048	0.804	1.365	0.234 ** (3.952)	1.263		1.003	1.591
ACMEETINGS	0.003 (0.002)	1.003	0.893	1.126	0.162 *** (9.846)	1.176		1.063	1.302
ACINDEPENDENCE	3.243 *** (6.924)	25.606	2.287	286.661	−1.172 * (3.810)	0.310		0.096	1.005
Constant	−14.810 *** (33.728)	0.000			−1.394 (0.534)	0.248
Years Effect	Included				Included
Industries Effect	Included				Included
Observations	612				612

* is significant at level ˂ 10%, ** is significant at level ˂ 5%, *** is significant at level ˂ 1%.

, panels A, B, and C. The omnibus test results presented in panel A show that the ESG affects the AC and AO, as the chi-squared value of the omnibus test is significant at the 1% level for both models. This indicates that the independent variables in our models have a significant effect on both AC and AO. Thus, the BLR models are fitted to predict the relation between the dependent and independent variables.

To ensure that the models are fitted with the observation data, we depend on the Hosmer–Lemeshow tests which are required in BLR analysis. Panel B reveals chi-squared values of 9.455 and 11.310 with a significance of 0.305 and 0.185 for models (1) and (2), respectively. These values are greater than the alpha value of 0.05, indicating that the BLR models are fitted with our data, and, hence, our models can predict the relation between ESG, AC, and AO from the existing data.

The ability of ESG, in total, to illustrate the variances in AC and AO variables in the BLR models was tested using Nagelkerke R² presented in Panel C. The Nagelkerke R² values are 0.464, and 0.384 for models (1) and (2), respectively, indicating that 46.4% of the variance in AC and 38.4% of the variance in AO could be explained by ESG or independent variables in our models. Overall, the results show that BLR models can significantly predict the relation between ESG, AC, and AO.

Panel D confirms the effect of independent variables individually on AC and AO. The coefficient value of ESG in model (1) is positive (1.127) and significant at the 1% level, meaning that the ESG firms tend to assign one of the Big4 auditors. Therefore, we accept H1. This result is consistent with the view that ESG firms are more likely to demand higher audit quality to enhance financial reporting quality [65,66,67]. Model (1) also presents the probability of ESG firms assigning Big4 auditors. The Exp (B) for firms indexed in the ESG index is 3.088. This finding implies that the likelihood of ESG firms assigning one of the Big4 auditors is 3.088 times higher than that of other firms. However, the coefficient value of the ESG in model (2) is negative (−1.083) and significant at the 1% level. This finding means that ESG firms are less likely to receive a qualified audit opinion or they are more likely to receive an unqualified audit opinion. This result ensures that a sound ESG performance improves financial reporting quality, which consequently reduces the likelihood of issuing a qualified opinion [59,64,65]. Therefore, we accept H2. Model (2) also presents the probability of ESG firms receiving a qualified audit opinion. The Exp (B) for firms indexed in the ESG index is 0.339. This finding suggests that ESG firms are 0.661 times less likely to receive a qualified audit opinion than other firms.

4.3. Robustness Tests

To ensure the robustness of our main results, we first addressed the concerns of the potential simultaneity by re-estimating models (1) and (2) by including a one-year lag for independent and control variables [32]. By doing so, we allow time for the effects of ESG on AC and AO to be discerned. The results reported in

Table 6. Robustness tests using lagged ESG.

Panel A: Omnibus Test of Model Coefficients
Model (1): Auditor Choice					Model (2): Audit Opinion
Step 1	Chi-square	D.f.	Sig.		Chi-square	D.f.	Sig.
Step	240.272	24	0.000		205.629	24	0.000
Block	240.272	24	0.000		205.629	24	0.000
Model	240.272	24	0.000		205.629	24	0.000
Panel B: Hosmer–Lemeshow’s goodness-of-fit test
Model (1): Auditor Choice					Model (2): Audit opinion
	Chi-square	D.f.	Sig.		Chi-square	D.f.	Sig.
Step 1	10.690	8	0.220		9.134	8	0.331
Panel C: Nagelkerke R2 square test
Model (1): Auditor Choice					Model (2): Audit opinion
	−2 Log likelihood	Cox and Snell R2	Nagelkerke R2		−2 Log likelihood	Cox and Snell R2	Nagelkerke R2
Step 1	533.158	0.325	0.453		624.330	0.285	0.384
Panel D: Wald test
Variables	Model (1): Auditor Choice				Model (2): Audit opinion
	B (Wald)	Exp (B)	95% C.I. for Exp (B)		B (Wald)	Exp (B)	95% C.I. for Exp (B)
			Lower	Upper			Lower	Upper
Step 1 Lagged ESG	0.829 *** (9.348)	2.292	1.347	3.899	−1.102 *** (14.394)	0.332	0.188	0.587
FSIZE	0.622 *** (46.301)	1.863	1.557	2.229	0.003 (0.001)	1.003	0.859	1.171
LEVERAGE	0.242 (0.179)	1.274	0.416	3.904	0.499 (0.904)	1.648	0.589	4.612
PROFITABILITY	0.222 (0.019)	1.249	0.053	29.531	0.379 (0.066)	1.461	0.081	26.356
FGROWTH	0.064 (0.062)	1.066	0.645	1.762	−0.485 ** (4.123)	0.616	0.386	0.983
LOSS	−0.145 (0.181)	0.865	0.443	1.689	1.340 *** (20.478)	3.820	2.138	6.826
FAGE	−1.073 *** (22.910)	0.342	0.220	0.531	0.444 ** (4.700)	1.559	1.044	2.330
BSIZE	−0.112 ** (5.381)	0.894	0.813	0.983	−0.036 (0.593)	0.964	0.879	1.058
BMEETINGS	−0.095 *** (11.821)	0.910	0.862	0.960	0.095 *** (14.699)	1.099	1.047	1.154
DUALITY	−0.554 ** (4.775)	0.574	0.349	0.944	0.228 (0.852)	1.256	0.774	2.040
BINDEPENDENCE	2.533 *** (10.837)	12.596	2.787	56.921	−2.555 *** (15.378)	0.078	0.022	0.279
ACSIZE	0.057 (0.176)	1.058	0.812	1.380	0.233 ** (3.909)	1.262	1.002	1.590
ACMEETINGS	0.008 (0.020)	1.008	0.899	1.131	0.162 *** (9.716)	1.176	1.062	1.301
ACINDEPENDENCE	3.467 *** (7.505)	32.055	2.683	383.043	−1.155 * (3.699)	0.315	0.097	1.022
Constant	−15.078 *** (34.490)	0.000			−1.396 (0.539)	0.248
Years Effect	Included				Included
Industries Effect	Included				Included
Observations	612				612

* is significant at level ˂ 10%, ** is significant at level ˂ 5%, *** is significant at level ˂ 1%.

are consistent with those presented in Table 5. Panel D shows that the coefficient value of the lagged ESG in model (1) is positive (0.829) and significant at the 1% level, meaning that the ESG firms tend to assign one of the Big4 auditors in the next year. The Exp (B) for firms indexed in the ESG index is 2.292, suggesting that the ESG firms are 2.292 times more likely to assign one of the Big4 auditors than other firms. However, the coefficient value of ESG in model (2) is negative (−1.102), and significant at the 1% level, meaning that the ESG firms are less likely to receive a qualified audit opinion in the subsequent year. Model (2) also presents the probability of ESG firms receiving a qualified audit opinion, where the Exp (B) for firms indexed in the ESG index is 0.332. This finding implies that ESG firms are 0.668 times less likely to receive a qualified audit opinion in the subsequent year than other firms.

We addressed the potential endogeneity problems using instrumental variables (IVs). We followed previous research by employing two IVs to predict ESG in addition to control variables (e.g., [57,68]). The first IV is the average ESG in the industry because the firm-level ESG is closely related to industry norms, and the second IV is two years lagged ESG at the firm level. The results reported in

Table 7. Robustness tests using predicted ESG.

Panel A: Omnibus Test of Model Coefficients
Model (1): Auditor Choice					Model (2): Audit Opinion
Step 1	Chi-Square	D.f.	Sig.		Chi-Square	D.f.	Sig.
Step	234.705	24	0.000		200.570	24	0.000
Block	234.705	24	0.000		200.570	24	0.000
Model	234.705	24	0.000		200.570	24	0.000
Panel B: Hosmer–Lemeshow’s goodness-of-fit test
Model (1): Auditor Choice					Model (2): Audit opinion
	Chi-square	D.f.	Sig.		Chi-square	D.f.	Sig.
Step 1	14.397	8	0.072		14.507	8	0.069
Panel C: Nagelkerke R2 square test
Model (1): Auditor Choice					Model (2): Audit opinion
	−2 Log likelihood	Cox and Snell R2	Nagelkerke R2		−2 Log likelihood	Cox and Snell R2	Nagelkerke R2
Step 1	538.725	0.319	0.444		629.390	0.279	0.376
Panel D: Wald test
Variables	Model (1): Auditor Choice				Model (2): Audit opinion
	B (Wald)	Exp (B)	95% C.I. for Exp (B)		B (Wald)	Exp (B)	95% C.I. for Exp (B)
			Lower	Upper			Lower	Upper
Step1 Predicted ESG	0.555 ** (3.863)	1.741	1.002	3.027	−0.942 *** (9.836)	0.390	0.216	0.702
FSIZE	0.642 *** (48.771)	1.899	1.586	2.274	−0.001 (0.000)	0.999	0.855	1.168
LEVERAGE	0.202 (0.127)	1.223	0.404	3.700	0.503 (0.927)	1.653	0.594	4.602
PROFITABILITY	0.200 (0.015)	1.221	0.051	29.326	0.525 (0.129)	1.691	0.096	29.644
FGROWTH	0.061 (0.056)	1.062	0.644	1.753	−0.487 ** (4.166)	0.615	0.385	0.981
LOSS	−0.193 (0.326)	0.825	0.425	1.599	1.381 *** (21.948)	3.978	2.233	7.088
FAGE	−1.122 *** (24.906)	0.326	0.210	0.506	0.456 ** (4.959)	1.577	1.056	2.356
BSIZE	−0.111 ** (5.162)	0.895	0.812	0.985	−0.036 (0.574)	0.965	0.880	1.058
BMEETINGS	−0.096 *** (12.270)	0.908	0.861	0.959	0.095 *** (15.012)	1.099	1.048	1.153
DUALITY	−0.560 ** (4.876)	0.571	0.347	0.939	0.204 (0.676)	1.226	0.754	1.996
BINDEPENDENCE	2.524 *** (10.717)	12.473	2.753	56.510	−2.588 *** (15.854)	0.075	0.021	0.269
ACSIZE	0.069 (0.257)	1.071	0.821	1.398	0.231 ** (3.861)	1.260	1.001	1.586
ACMEETINGS	0.010 (0.029)	1.010	0.901	1.133	0.160 *** (9.637)	1.174	1.061	1.299
ACINDEPENDENCE	3.693 *** (8.151)	40.145	3.182	506.489	−1.217 ** (4.124)	0.296	0.091	0.958
Constant	−15.409 *** (35.122)	0.000			−1.364 (0.507)	0.256
Years Effect	Included				Included
Industries Effect	Included				Included
Observations	612				612

** is significant at level ˂ 5%, *** is significant at level ˂ 1%.

are consistent with those presented in Table 5. Panel D reveals that the coefficient value of the predicted ESG in model (1) is positive (0.555) and significant at the 5% level, meaning that the ESG firms tend to assign one of the Big4 auditors. The Exp (B) for firms indexed in the ESG index is 1.741, indicating that the ESG firms are 1.741 times more likely to assign one of the Big4 auditors than other firms. However, the coefficient value of the ESG in model (2) is negative (−0.942) and significant at the 1% level, meaning that the ESG firms are less likely to receive a qualified audit opinion. Model (2) also presents the probability of ESG firms receiving a qualified audit opinion, where the Exp (B) for firms indexed in the ESG index is 0.390. This ensures that the ESG firms are 0.61 times less likely to receive a qualified audit opinion than other firms.

We re-ran the models depending on logistic regression using cluster robust standard error method, where observations are clustered by firm. This method leads to significantly more accurate inference in finance panels [69]. The results in

Table 8. Logistic regression using the cluster robust standard error method.

	Model (1): Auditor Choice	Model (2): Audit Opinion
ESG	1.13 *** (4.18)	−1.08 *** (−3.87)
FSize	0.60 *** (7.11)	0.0046 (0.060)
Leverage	0.28 (0.46)	0.51 (1.03)
Profitability	−0.11 (−0.082)	0.64 (0.45)
FGrowth	0.050 (0.21)	−0.50 ** (−2.17)
LOSS	−0.15 (−0.48)	1.37 *** (4.40)
FAge	−1.00 *** (−4.50)	0.44 ** (2.12)
BSIZE	−0.12 *** (−2.69)	−0.036 (−0.80)
BMEETINGS	−0.092 *** (−3.25)	0.094 *** (3.88)
DUALITY	−0.53 ** (−2.04)	0.230 (0.92)
BINDEPENDENCE	2.71 *** (3.63)	−2.59 *** (−4.00)
ACSIZE	0.047 (0.37)	0.230 ** (2.04)
ACMEETINGS	0.0027 (0.053)	0.16 *** (3.34)
ACINDEPENDENCE	3.24 *** (3.53)	−1.17 ** (−2.00)
Constant	−13.0 *** (−6.44)	−0.97 (−0.51)
Years Effect	Included	Included
Industries Effect	Included	Included
N	594	612
Pseudo R2	0.3075	0.2474

** is significant at level ˂ 5%, *** is significant at level ˂ 1%.

are consistent with those presented in Table 5, Table 6 and Table 7. The results confirm the view that ESG firms are more likely to assign one of the Big4 auditors and less likely to receive a qualified opinion.

4.4. Additional Analysis: ESG Performance, Auditor Choice, and Audit Opinion during COVID-19

Previous literature refers to the implications of COVID-19 on financial reporting quality and auditing outcomes [70]. For further insights in this regard, we examined the probable effect of COVID-19 on our hypotheses by adding COVID-19 as a moderator in our models. The results shown in

Table 9. Testing the research hypotheses during COVID-19.

Panel A: Omnibus Test of Model Coefficients
Model (1): Auditor Choice					Model (2): Audit Opinion
Step 1	Chi-Square	D.f.	Sig.		Chi-Square	D.f.	Sig.
Step	249.064	18	0.000		204.980	18	0.000
Block	249.064	18	0.000		204.980	18	0.000
Model	249.064	18	0.000		204.980	18	0.000
Panel B: Hosmer–Lemeshow’s goodness-of-fit test
Model (1): Auditor Choice					Model (2): Audit opinion
	Chi-square	D.f.	Sig.		Chi-square	D.f.	Sig.
Step 1	11.073	8	0.198		7.290	8	0.506
Panel C: Nagelkerke R2 square test
Model (1): Auditor Choice					Model (2): Audit opinion
	−2 Log likelihood	Cox and Snell R2	Nagelkerke R2		−2 Log likelihood	Cox and Snell R2	Nagelkerke R2
Step 1	524.366	0.334	0.466		624.980	0.285	0.383
Panel D: Wald test
Variables	Model (1): Auditor Choice				Model (2): Audit opinion
	B (Wald)	Exp (B)	95% C.I. for Exp (B)		B (Wald)	Exp (B)	95% C.I. for Exp (B)
			Lower	Upper			Lower	Upper
Step 1 ESG	0.830 *** (6.858)	2.292	1.232	4.265	−1.483 *** (16.263)	0.227	0.110	0.467
COVID-19	−1.634 *** (22.300)	0.195	0.099	0.384	−0.078 (0.107)	0.925	0.580	1.476
COVID-19*ESG	1.027 ** (3.612)	2.793	0.968	8.054	1.055 ** (4.050)	2.871	1.028	8.017
FSIZE	0.586 *** (41.363)	1.797	1.503	2.148	−0.019 (0.061)	0.981	0.841	1.144
LEVERAGE	0.277 (0.228)	1.319	0.423	4.113	0.448 (0.741)	1.566	0.564	4.348
PROFITABILITY	−0.255 (0.026)	0.775	0.035	17.348	0.198 (0.019)	1.219	0.074	20.176
FGROWTH	−0.004 (0.000)	0.996	0.607	1.633	−0.510 ** (4.840)	0.601	0.381	0.946
LOSS	−0.140 (0.168)	0.869	0.445	1.699	1.342 *** (20.901)	3.825	2.152	6.799
FAGE	−1.014 *** (20.568)	0.363	0.234	0.562	0.401 ** (3.922)	1.493	1.004	2.220
BSIZE	−0.114 ** (5.604)	0.892	0.812	0.981	−0.028 (0.353)	0.973	0.887	1.066
BMEETINGS	−0.089 *** (10.523)	0.914	0.866	0.965	0.097 *** (15.360)	1.101	1.049	1.156
DUALITY	−0.569 ** (4.915)	0.566	0.342	0.936	0.187 (0.573)	1.206	0.743	1.958
BINDEPENDENCE	2.675 *** (11.825)	14.508	3.159	66.632	−2.668 *** (16.678)	0.069	0.019	0.250
ACSIZE	0.025 (0.036)	1.026	0.789	1.333	0.225 * (3.638)	1.252	0.994	1.578
ACMEETINGS	0.013 (0.048)	1.013	0.903	1.137	0.171 *** (10.665)	1.186	1.071	1.314
ACINDEPENDENCE	3.215 *** (6.810)	24.908	2.226	278.652	−1.176 ** (3.866)	0.308	0.095	0.996
Constant	−12.787 *** (28.086)	0.000			−0.592 (0.106)	0.553
Years Effect	Not included				Not included
Industries Effect	Included				Included
Observations	612				612

* is significant at level ˂ 10%, ** is significant at level ˂ 5%, *** is significant at level ˂ 1%.

are consistent with those presented in Table 5, suggesting that COVID-19 has a significant negative effect on AC (at the 1% level) and an insignificant negative effect on AO. The results also reveal an increase in the likelihood of ESG firms assigning one of the Big4 auditors during COVID-19 as the coefficient value of COVID-19*ESG is 1.027 in model (1), and this result is significant at the 5% level. However, the results suggest an increase in the likelihood of ESG firms receiving a qualified audit opinion during COVID-19, as the coefficient value of COVID-19*ESG is 1.055 in model (2), and this result is significant at the 5% level. These results are consistent with Hsu and Yang, who found a decrease in UK companies’ financial reporting quality during the pandemic [70].

5. Conclusions

This study has examined the relationship between ESG performance, auditor choice, and audit opinion. There are two arguments in this regard. The first one indicates that firms with higher ESG performance will be more ethical, and motivated to demand higher audit quality through assigning one of the Big4 auditors to provide transparent information to the stakeholders. On the contrary, firms may engage in ESG activities to conceal their misbehaviors. Consequently, the real intention of ESG performance, under this argument, is to mislead stakeholders with opportunistic behaviors, negatively influencing financial reporting. Then, ESG firms are less likely to demand higher audit quality. In addition, we examined the relation between ESG performance and audit opinion. To test our hypotheses, we used a sample of listed firms on EGX100 during the period 2014–2022. The results revealed that higher ESG performance firms are more likely to assign one of the Big4 auditors. These results are consistent with the view that firms with higher ESG performance are more likely to demand higher audit quality to enhance financial reporting quality [65,67]. In addition, our results indicated that higher ESG performance firms are more likely to have an unqualified audit opinion. This is consistent with the view that auditors in these firms are less likely to issue a qualified opinion, which ensures that ESG performance improves financial reporting quality [59,64,65]. Moreover, as an additional analysis, we examined the effect of COVID-19 on our results. The results showed an increase in the likelihood of ESG firms assigning one of the Big4 auditors during COVID-19. However, COVID-19 increases the likelihood of ESG firms receiving a qualified audit opinion, which is consistent with Hsu and Yang who reported a decrease in financial reporting quality during the pandemic period [70].

Our study contributes to the literature as the first study that examines the relationships between ESG, AC, and AO in Egypt and considers the effect of COVID-19 on these relationships. Our results support the ethical perspective of ESG firms in Egypt. The current findings provide significant evidence to policymakers, auditors, and investors in emerging markets. They can guide policymakers in formulating guidelines and regulations to better oversee firms, enhance governance, and protect stakeholders’ interests, especially during crises. Further, the current findings advise investors to consider ESG performance while making investment decisions, especially during crisis time.

However, this study is not without limitations. Some variables that may affect ESG performance, such as ownership concentration, political connections, and institutional ownership, were not examined in our study. Future research may consider these governance variables for new insights concerning the ESG–audit practices relationship. Considering the focus of this study on the Egyptian market, we suggest that future research could reinvestigate the relationship between ESG performance, AC, and AO in other countries with different cultural and institutional contexts. Further, employing qualitative research methods such as case studies and interviews in future research may add further insights concerning the implications of ESG for the auditors’ decisions during different audit phases.