Corporate Social Responsibility Risk and Audit Opinion

This article studies the impact of corporate social responsibility on audit costs. Most of the previous studies are based on the social responsibility report level. This paper starts with the third party rating agency’s assessment of corporate social responsibility risk to carry out research on its impact on audit costs. It is found that corporate social responsibility risk will significantly increase the audit costs. At the same time, this paper finds that the audit of the company by the international big four accounting firms or the company’s social responsibility report may reduce the impact of social responsibility risk on audit costs.


Introduction
Social responsibility was first proposed by Sheldon in 1924. Since then, with the global emphasis on environmental and human rights issues, more and more scholars have launched research on corporate social responsibility. Investors start a business with the intention of making a profit. But in order to become a strong player in social competition as soon as possible, some of them excessively pursue and extremely exploit stakeholders, which leads to environmental degradation, harsh treatment of employees, and decline in product quality. Due to the continuous oppression of workers and the increasing demands of the working class to safeguard their rights and interests, the corporate social responsibility movement in the 1980s began to rise in developed countries in Europe and the United States. With the successive promulgation of relevant international treaties, countries attach great importance to social responsibility and gradually improve the system of corporate social responsibility performance, and the concept of social respon-sibility has become an international consensus [1]. Reputation is an important asset of a company [2] [3]. Today, companies that want to guarantee their survival must bear social obligations and the relevant social costs, and it plays an active role in improving society. Based on this, this article conducts research on social responsibility in the context of the Chinese market.
This article selects the Chinese market because China's legal development in social responsibility started late and is currently being continuously improved. The rapid development of China's economy in recent years has also exposed more and more problems, and social responsibility scandals have frequently appeared. Recently there was a serious vaccine scandal which significantly affects the company's image. And the market value of the listed company had shrunk rapidly and finally was forced to delist due to the incident. Behind this is the lack of credit of the company's management team. This incident triggered a serious discussion of corporate social responsibility from the whole society once again. China is a developing country. Since the reform and opening-up, China's economy has undergone tremendous changes, but we must also be aware of some negative factors behind this rapid development. In the decades of China's rapid economic development, environmental pollution incidents have occurred frequently. There have been more and more incidents of squeezing employees, wage arrears, and even death from overwork. Incidents such as tax evasion and money laundering have been common occurrences, and product problems have continuously threatened our lives. This is inevitable for economic development to a certain level. Therefore, the current Chinese market is more suitable for research on this relationship. With the progress of Chinese society and the increasing awareness of national self-protection, the implementation of corporate social responsibility has attracted more and more attention from the society. Since January 1, 2006, the provisions of Article 5 of the Chinese Company Law on the implementation of corporate social responsibility have come into effect. For today's fierce competition between enterprises, if enterprises want to continue to develop and occupy a place, they should not only be responsible for profit, but also responsible for the environment, society and the country, bear the corresponding social responsibility, and establish a good image.
As an important intermediary agency in the capital market, an accounting firm can provide external investors with relevant information about the reliability of the company's financial reports and reduce the impact of information asymmetry. When auditing an enterprise, the certified public accountant will review the authenticity and accuracy of its financial data and judge the effectiveness of internal control, and issue an audit opinion. The audit opinion issued by it will have a significant impact on the future operation and development of the enterprise. China's 2006 auditing standard system established a risk-based auditing model. This standard increases the legal risks of auditors, and auditors need to take more responsibility for their audit opinions. In the face of high corporate risk, auditors will be more careful in reviewing the company before issuing an audit opinion, which will increase audit costs [4].

Y. Fang American Journal of Industrial and Business Management
At present, there are few studies on the relationship between social responsibility and audit opinions. And most of them are based on whether corporate establish a social responsibility report or the degree of disclosure of the report. Scholars have found that the positive performance of social responsibility is negatively related to the possibility that companies are issued non-standard audit opinions.
For example, Song Zhiqiong and Wang Xiaojun [5] found that disclosure of corporate social responsibility reports will reduce the possibility of non-standard audit opinions. Social responsibility but corporate self-disclosure is often lack of objectivity. Wang Juan and Pan Xiuli [6] found that charitable donations would reduce the possibility of being issued non-standard audit opinions. But they also found that companies with higher levels of earnings management would use charitable donations to cover up corporate misconduct. This is consistent with the suspicion of "goodwill" discovered by Quan Xiaofeng, Wu Shinong and Yin Hongying [7]. This further shows the credibility of corporate goodwill social responsibility behavior. As a result, the market's response to the company's active performance of social responsibility will gradually weaken. Therefore, in order to make up for the shortcomings of the existing research, this article will start from a new perspective, namely, using social responsibility risk to carry out research on the relationship between the two. The research in this paper enriches the existing research on the influencing factors of audit opinions and the impact of social responsibility risk on enterprises.
The audit opinion passed the information of the authenticity and reliability of the company's financial report to external investors, which has an important impact on the company's future operations, and the performance of social responsibility has also received increasing attention. Therefore, it is of great significance for the company to verify whether social responsibility risks will affect audit opinions through empirical research. It is hoped that through the research in this article, enterprises will be more actively fulfilled in their social responsibilities, so as to promote the healthy and stable development of China's economy and make people's lives happier.
Based on the research of Chinese and foreign scholars and the information asymmetry theory, halo effect, and signal transmission theory, three hypotheses were proposed. Taking the listed companies in China from 2011 to 2016 as the research sample (the financial industry companies have been deleted because the financial indicators of financial companies are significantly different from other companies), based on the corporate social responsibility risk, I study the listed companies' social responsibility on audit opinions. Through empirical research, it is verified that social responsibility risk will affect the audit opinions issued by auditors. It proves that the auditor will identify the social responsibility risk and issue a more objective audit opinion. This impact will vary depending on whether the Big Four accounting firms issued audit opinions. In addition, whether a social responsibility report is issued affect the relationship between the two.

Literature Review
The Impact of Social Responsibility on Companies The research on social responsibility started early in foreign countries and can be traced back to the 1920s. The related research results are also very rich. Early scholars, such as Bragdon and Marlin [8] and Monskowitz [9], discovered that the positive relationship between corporate social responsibility and corporate financial performance. Regarding the mechanism, Li Guoping, Zhang Qianqian and Zhou Hong [10] explained the positive and negative impact mechanism by summarizing the foreign conclusions on the impact of corporate social responsibility on financial performance. They gave four explanations for the positive correlation mechanism: reducing capital costs, gaining competitive advantages, gaining reputation and other intangible assets, and risk management. Schnietz [11] found that during the period of corporate crisis, corporate social responsibility surplus can effectively prevent the company's stock price from falling. This also reflects the risk management mechanism of social responsibility, showing that social responsibility performance can effectively divert public attention and cover up other improper behaviors of enterprises. In addition, Li Shu and Xie Xiaoyan [12] found that the higher the performance of private corporate social responsibility, the more helpful it is to obtain loans. But Alexander & Buchholz used stock returns as a performance measure and found a negative correlation between the two. But most scholars' research still supports the positive correlation between the two [13]. Byun and Oh [14] found that public social responsibility activities are positively related to shareholder value, improving future business performance. This has also led to news buying behavior by many companies.
In view of this, some scholars inevitably questioned the "goodwill" of enterprises. Quan Xiaofeng, Wu Shinong and Yin Hongying [7] questioned the behavior of enterprises in fulfilling their social responsibilities. They verified the characteristics of the social responsibility as a self-interest tool through research on Chinese listed companies.
Although the fulfillment of social responsibility can bring many positive effects to the enterprise, it will also increase the agency cost of the enterprise and occupy enterprise's resources. Compared with individual investors, institutional investors are often more rational and more profitable. Fernando [19] found that good corporate governance quality is more susceptible to standard unqualified audit opinions. Lv Minkang and Liu Zheng [20] found that companies with higher investor attention are more likely to receive non-standard audit opinions. This is a self-protection phenomenon due to the higher cost of failure faced by auditors. Fang, Gerald, Zhang and Zhao [21] found that the companies with more related party transactions are more likely to give non-standard audit opinions. Moreover, they point out that the way that auditors reduce the audit risk brought by related party transaction risks is to provide more cautious audit opinions, instead of increasing audit efforts.
In addition to the characteristics of audit objects, audit subjects also affect the types of audit opinions. Many scholars have found that the larger the size of the accounting firm, the higher the probability of issuing non-plot audit opinions.
Karjalainen, Niskanen and Niskanen [22] studied the influence of audit partner gender on audit opinions based on the auditor's personal characteristics. They found that female audit partners are more inclined to issue non-standard audit opinions after the audit partner changes.
Chen, Zhang and Zhou [23] studied the impact of national culture on audit opinions at a macro level. They found that auditors are more likely to give non-standard audit opinions to customers in countries with a stronger culture of confidentiality, and stronger investor protection would weaken the relationship between the two.

Research on the Impact of Social Responsibility on Audit Opinions
There is very little research on the impact of social responsibility on audit opinions. Wang Juan and Pan Xiuli [6] have found that charitable donations can reduce the possibility of companies being issued non-standard audit opinions.
And listed companies with higher earnings management are more motivated to Social responsibility reports often lack objectivity. And in the face of more and more social responsibility performance acts as "tools" and news purchase behavior, positive social responsibility performance is no longer "purity", and the outside world has begun to realize that this social responsibility performance can no longer be regarded as it is a reflection of the low level of corporate earnings management. Therefore, the impact of social responsibility on audit opinions needs to be addressed from a more effective perspective.

Information Asymmetry Theory
As an important intermediary agency in the capital market, accounting firms will review the authenticity and accuracy of corporate financial data and judge the effectiveness of internal control, and issue audit opinions. As an external monitoring mechanism, accounting firms can provide information to external investors and reduce the impact of information asymmetry. But it is also a third-party organization that can be also affected by information asymmetry. Chen, Zhang and Zhou [23] found that auditors are more likely to give non-standard audit opinions to customers in countries with a stronger culture of confidentiality. It can be seen that, for those actions that significantly increase the asymmetry of information, auditors are more likely to issue non-standard audit opinions for self-protection. And reducing information asymmetry will reduce the possibility of being issued non-standard audit opinions [5]. Fang, Gerald and Zhang [21] found that the companies with more related party transactions are more likely to give non-standard audit opinions. Liu Xiaolun, Hao Chen and Chu Yuping [19] found that good corporate governance quality is more susceptible to standard unqualified audit opinions. It can be seen that the signal to the outside world about the degree of earnings management has a positive correlation with the non-standard audit opinion. Similar to these factors, in the case of asymmetric information, the higher risk of corporate social responsibility often sends a signal to the outside world that the degree of corporate earnings management is high.

Halo Effect
Wang Juan and Pan Xiuli [6] have pointed out that charitable donations can alleviate the possibility of non-standard audits for companies with higher earnings management. It can be seen that when a company actively promotes the performance of social responsibility, it is often labeled as "good" by the outside world because of these positive behaviors of the enterprise, and is thought to have "good quality" for everything. If a company's social responsibility risk is high, it will be considered as a "bad" company, and all its actions will be ques- Due to the existence of information asymmetry, auditors will be affected by the theory of signal transmission and halo effect in companies facing higher CSR risks, and they will have the impression that the company has a higher degree of earnings management. Too much attention to social responsibility risks, in order to reduce their own audit risks, auditors may reduce their own risks by issuing more cautious audit opinions. As a result, Hypothesis 1 is proposed.
H1: Corporate social responsibility risk is negatively related to audit opinions, that is, the higher the corporate social responsibility risk, the easier it is to obtain non-standard audit opinions.
In general, the auditors of the Big Four accounting firms have stronger professionalism, so it is inferred that they can overcome psychological deviations and issue more objective audit opinions when facing corporate social responsi- H2: Whether audited by big four has no effect on the relationship between corporate social responsibility risk and audit opinions.
Issuing a social responsibility report can reduce the asymmetry of information, and it also sends a signal to the outside that the company is more actively performing social responsibility, which can alleviate the impression deviation caused by the halo effect to auditors in some extent. This leads to Hypothesis 3.
H3: If the company issue social responsibility reports, the negative impact of corporate social responsibility risks on audit opinions will be weaker.

Independent Variable: Social Responsibility Risk
Most previous research has used the presence or disclosure of corporate social responsibility reports to measure corporate social responsibility. But with the company's social responsibility performance is perceived as self-profit tool, often the company's negative information is more valuable and more likely to attract external attention. Therefore, in order to improve the objectivity and effectiveness of social responsibility indicators, this article uses the Reprisk database, a third-party organization's measure of corporate social responsibility risk. The social responsibility risk of the RepRisk database collects corporates' negative news on environment, society and governance. I use two variables to measure reputational risk. They are risk rating and the number of negative news. The Re-pRisk database divides risks from low to high into ten levels: AAA, A, A, BBB, BB, B, CCC, CC, C, and D, which are converted to numbers 1-10 in this paper.

Dependent and Regulated Variables
Dependent variable in this paper is audit opinion (AuditOpinion). If the company received a non-standard audit opinion, it takes 1, otherwise 0.

Model Design
I proposed the regression model of this article based on previous research.

Main Regression Model:
, 0 As for the calculation of earnings quality, I use the modified Jones model.

Variable Definition
By studying previous research, I summarized the variables in this article and explained how they are calculated.
As shown in Table 1, the calculation of most variables follows the former researchers.

Empirical Analysis
Taking the data of listed companies in Shanghai and Shenzhen in 2011 and 2016 as samples, descriptive statistics and multiple regression analysis were used to verify whether social responsibility risks would affect the audit opinions issued by auditors.

Descriptive Statistics
In order to better understand the data, this paper makes descriptive statistics on the relevant data to show the general situation of the sample. Table 2 shows the descriptive statistics of main variables for the full sample. In Table 2, we can see that only 5.3% of the observations received non-standard opinion. And the average corporate rating is 5.431, which is above medium risk.

Y. Fang American Journal of Industrial and Business Management
As for news, there were less than two pieces of news on average for each company per year. But the max value reaches to 254, which is a very high value. In another words there is a piece of news per 1.4 days. We can see from the table that there are nearly 13% observations are audited by 4 international big companies, and they have good market influence. There are 46.6% of observations disclose responsibility report.
In addition to the descriptive statistics of the whole sample, this paper also statistics relevant data based on the type of audit opinion. This will help us understand the differences between the two samples in advance. Table 3 shows sub-samples grouped by audit opinion. We can see that Ri-skRating of observations which received non-standard opinion is 5.610 on average, and 5.421 for standard opinion group. Based on t-test, it can be found that the samples of non-standard audit opinions are issued with significantly higher social responsibility risks, which indicates that companies with higher social responsibility risks are more likely to obtain non-standard audit opinions. The sample of the standard audit opinion issued has a lower degree of earnings management. At the same time, it can be found that, compared with non-standard samples, the samples audited with standard audit opinions have better profitability; the probability of the standard audit opinions issued by the Big Four is higher than that of other accounting firms, which may indicate that the quality of the Big Four clients is higher; And audit costs will be higher.
Except for descriptive statistics, I also study these variables' correlations to check the hypotheses basically.  In Table 4, we can see that there is a significant positive correlation between social responsible risk and audit opinion at 1% significant level, which is consistent with my hypothesis. But news is not significantly related to Audit Opinion.
There is also a significant positive correlation between |DA| and audit opinion, which is also consistent with previous studies. Lev has a significant positive correlation with audit opinion, and cash flow has a significant negative correlation with audit opinion, indicating that companies with greater financial risk are more vulnerable. The correlation of other variables is basically consistent with expectations.

Regression Analysis
To test the hypotheses, I use Stata to perform regression analysis. Table 5 shows the test results of Hypothesis 1.
From the logit regression in Table 5, the coefficients of social responsibility risk are both significantly positive, both in terms of social responsibility risk level and the number of social responsibility negative news (both pass the significance level of 0.01). For every additional unit of RiskRating, the possibility of being audited with an unqualified opinion will increase by 0.34%. As for News, the coefficient before it is 0.0158. This shows that as corporate social responsibility risks increase, auditors are more likely to issue non-standard audit opinions.
Consistent with the forecast, financial leverage has a significant positive correlation with audit cost and the coefficients before them are 2.7035 and 2.6793. We can also see that four international big audit companies are more tend to give a Y. Fang qualified audit opinion. This is significant at 1% level with a number of −1.2461% in column 1. Most of the correlations are consist with the expectation.
At the same time, I found that Rating would significantly increase the cost of the audit, but it did not significantly reduce the impact of Rating on AuditOpinion in the unlisted regression. There are two explanations for this: 1) Firms will increase audit fees due to the company's social responsibility risk to enhance their own returns, but will not increase audit procedures to identify whether these risks actually increase their own audit risks. 2) Auditors will increase audit procedures, and these risks do increase their own audit risks. But from the Pearson correlation coefficient, we can see that there is no significant linear relationship between earnings manipulation profit and social responsibility risk, which can eliminate the second explanation. Fang, Gerald and Zhang [21] have also pointed out that when faced with related party transactions, auditors mainly adjust their audit opinions rather than improve their auditing efforts.   To test the Hypotheses 2 and 3, I add crossover based on Table 5 and the result is shown in Table 6. From the logit regression of column (1), the coefficient before "RiskRating × Big4" is significantly negative at 1% significant level, indicating that the companies audited by the four major accounting firms will weaken the social responsible risk impact on their audit opinions that is decrease the possibility of being given an unqualified opinion by 1.0991%. The explanation for this may be that the auditors of the Big Four accounting firms are more able to objectively assess these risks, effectively overcome psychological deviations, and issue more effective audit opinions. From the logit regression of column (2), the coefficient before the "RiskRating × Social" is also significantly negative with the number of −0.5858. It may indicate that the company who issued the social responsibility report will weaken the impact of social responsibility risk on its audit opinion. The explanation for this may be that the auditor believes that actively disclosing the social responsibility report is a company's act of actively fulfilling its social responsibility, thereby reducing the impact of social responsibility risk to a certain extent.

Robustness Test
Some people may question the results, arguing that the significant relationship obtained by the regression results is due to the higher probability of earnings operations of companies with higher social responsibility risks, rather than the auditors' taking social responsibility risks into account. This interpretation can be denied from the following two evidences. The first evidence is the result from the Pearson correlation coefficient. As can be seen from the results in the table, there is no significant linear correlation between |DA| and RiskRating. The second evidence is the regression results. It can be seen that after adding |DA|, the coefficient before RiskRating is still significant. In other words, the impact of the social responsibility risk obtained in this paper on the audit opinion is determined by itself, not due to the impact of earnings quality.

Conclusions
The healthy development of the economy is especially important for China, which is in the process of transition, and is a good foundation for China to become a developed country. Based on the research of the existing literature and theoretical analysis, this paper proposes three hypotheses. Using the data of Chinese A-share listed companies, I study the relationship between social responsibility risk and audit opinions, and this relationship whether will be different between the four major and non-four major audited companies and whether they issued a social responsible report. The study found that: 1) Corporate social responsibility risk will affect the risk assessment of the certified public accountant. The certified public accountant will be more cautious, and more likely to issue non-standard audit opinions.
2) The impact of social responsibility risk on audit costs will vary depending on whether the accounting firm is one of the top four international firms. The relationship will be weaken if the companies audited by the four international firms.

Y. Fang American Journal of Industrial and Business Management
3) The impact of social responsibility risk on audit costs will vary depending on whether the company independently issues a social responsibility report. The audit opinions of enterprises that independently disclose social responsibility reports will be less affected by social responsibility risks.
It can be seen that social responsibility risk will affect the external evaluation of the company, and actively controlling social responsibility risk is of great significance to the company. Although the four major audits and the disclosure of social responsibility reports will reduce this effect, it still exists.
There are also some weaknesses in the paper. Firstly, I only study the observations from 2011 to 2016; some researchers may have concerns about sample size. As for the sample size, it's for the attainable of reputational risk data. But the sample isn't a small sample; the conclusion of the paper may not be significantly different. Secondly, due to the limited understanding of related fields, theoretical analysis may not be mature enough in this article. Finally, the possibility of being given an unqualified audit opinion for Chinese company is very low. And the impact of the social responsibility risk on audit opinion is only nearly 0.35%. So whether it will attract the attention of company is doubtable.
Social responsibility generally includes three aspects of environment, society and governance. This article does not decompose social responsibility, but considers it comprehensively. In the future, we can consider decomposing social responsibility and study which aspect of social responsibility risk is more sensitive by the outside world. You can also consider the impact of social responsibility risk on other aspects of the company's transaction costs, such as loans.