University of Amsterdam
Faculty Economics and Business
Master Accountancy & Control both variants
PAPER C
Does Corporate Social Responsibility influence
the earnings quality of a company?
Name:
Leon Hogenbirk
Student number:
10000761
Version:
12
thof August 2014
2 Abstract: This paper examines the relationship of CSR and earnings quality. Three different proxies of earnings quality are used to test this relationship. These proxies consist of restatements following from SEC-investigations, accounting restatements and material weaknesses (the
combination of SOX-302 and SOX-404 errors). This relationship is tested through examining if the CSR-score of companies has significant impact on predicting restatements. The samples for the three restatements consist of two samples, containing firm-year observations for U.S.
companies listed at the S&P 500 with a timeframe ranging from 2003 to 2009. The non-reliance restatements sample (restatements from the SEC and accounting restatements) contains 1533 observations and the material weakness sample contains 1225 observations. The results of this research show that internal control weaknesses and SOX-302 errors are more likely to happen for non-CSR companies. This suggests that CSR-firms tend to have better earnings quality because their internal control systems are of higher quality than non-CSR firms. No significant result was found, however for restatements from SEC-investigations and accounting restatements. This suggests that this difference in earnings quality is not enough to result in significant differences among non-CSR and CSR firms for non-reliance restatements, as those measures of deteriorating earnings quality are quite extreme.
3 Samenvatting: Dit onderzoek bestudeerde de relatie tussen CSR en de kwaliteit van de winst van een bedrijf. Drie verschillende meetwijzen zijn gebruikt om deze relatie te testen en die bestaan uit: SEC-onderzoeken, correcties van jaarverslag cijfers en materiële zwaktes (de combinatie van SOX-302 en SOX-404 zwaktes, ook wel interne controle zwaktes genoemd). Deze relatie is getest door te onderzoeken of de CSR-score van bedrijven een significante invloed heeft op de drie categorieën van correcties van jaarverslag cijfers. De datasets voor de drie meetwijzen bestaan uit twee groepen met observaties tussen 2003 en 2009. De dataset voor de non-reliance correcties (de combinatie van correcties die volgen uit SEC-onderzoeken en accounting
correcties) bestaat uit 1533 observaties en de dataset voor de materiele zwaktes bestaat uit 1225 observaties. De resultaten van dit onderzoek zijn dat de interne controle zwaktes en SOX-302 minder vaak voorkomen bij bedrijven met een focus voor CSR dan bij bedrijven die geen focus op CSR hebben. Geen significant resultaat was gevonden voor de correcties door
SEC-onderzoeken en accounting correcties. Dit suggereert dat het verschil in de kwaliteit van de winst dat veroorzaakt worden doordat de interne controles van een niet-CSR bedrijf dus vaker niet in orde zijn, niet genoeg is om in correcties van de jaarverslag cijfers te leiden.
4
Content
1. Introduction ... 5
2. Literary review and hypotheses ... 8
2.1 Corporate Social Responsibility ... 8
2.2 Earnings quality ... 14
2.3 CSR and earnings quality ... 17
2.4 Related literature ... 25
3. Research Design & Methodology ... 28
3.1 Sample and variable selection ... 28
3.2 Model selection ... 31
3.3 Sample creation ... 33
4. Results ... 34
4.1 Descriptive statistics ... 34
4.1.1 Non-reliance restatements ... 34
4.2.1 Non-reliance restatements sample ... 39
4.2.2 ICMW sample ... 43
4.3 Results regression analysis non-reliance restatements and ICMW ... 46
4.4 Sensitivity analysis and analysis of the results ... 51
5. Conclusion ... 53
Appendix A: List of KLD-variables used for this research. ... 56
Appendix B: Regression output sensitivity analysis and multicollinearity tests... 58
5
1.
Introduction
Corporate social responsibility is increasingly becoming more important to firms because of increasing demands of investors and customers who demand that a firm shows their responsibility towards social issues. Because of this growing concept, firms had to adopt social policies and had to show their commitment to social issues such as the environment, employees and acting
ethically and socially. Corporate social responsibility can be defined as the serious attempt to solve social problems caused wholly or in part by the corporation (Fitch, 1976, p. 38). Customer demands for social and ethical policies were increasing up to the point that firms could not ignore these demands and so CSR, became a universally accepted and promoted concept by
governmental, non-governmental and corporations in the late 1999’s (Lee, 2008). Because of this widespread concept, firms and managers had to show their commitment to social values and take up responsibility towards social issues.
This lead to new research which suggested that if managers are really more social, they should also be more likely to act more ethical in their other decisions such as their choices for reporting the financial numbers of the firm. These papers suggested that managers of CSR-firms could report higher earnings quality than other firms and so, translate this social commitment towards investors. This relationship can be explained by defining some concepts. Higher earnings quality can be defined as by Dechow et al. (2010), as providing more information about the features of a firm’s financial performance that are relevant for the making of a specific decision made by a specific decision-makers.
Earnings management could influence this earnings quality which is the attempt of a corporate officer to influence the short-term result of the firm (Sevin & Schroeder, 2005). The earnings quality of companies will decrease if the corporate officer would manage earnings for personal gains but it will increase if the manager reveals hidden information about the future prospects of the company through managing earnings. The suggestion made by this research such as from Chih et al. (2007) and Kim et al. (2012) is that managing earnings for personal gains is not really ethical and does not fit into the social policies of CSR which means that CSR-firms
6 would provide more information to investors about the performance of the company, through their financial numbers.
These theories from an ethical and political perspective, hypothesize that CSR and earnings quality could be positively related because those managers could have an increased moral incentive to do the right thing because of their adopted CSR-values (Caroll, 1979; Jones 1995). This would suggest that managers from CSR-firms would refrain from managing earnings for personal gains.
However, other papers suggest a different relationship. Prior et al. (2008) suggests that
companies with strong CSR values could have lower earnings quality because companies could try to disguise their poor quality of their financial statements through the adoption of social policies. The attention for the poor quality of earnings of the company, is deflected by showing more commitment to CSR so that managers can keep legitimizing themselves or the company to shareholders, stakeholders and society. The legitimacy theory is an example of a theory who explains this relationship through the fact that organizations continually seek to ensure that they operate within the bounds and norms of their societies. This can be done by showing desired behavior without actually behaving in the way that is shown to the public (Brown & Deegan, 1998, p.16).
These theories have been examined by a few papers who provide conflicting results. The
relationship between CSR and earnings quality such as Kim et al. (2012), Prior et al. (2008) and Chih et al.(2007) but those studies yielded different and inconclusive results. Kim et al. (2012) found, for example, that CSR could influence earnings quality positively by reducing earnings management by studying the discretionary accruals, real earnings management and the violations of GAAP. However, Prior et al. (2008) found that CSR increased the level of earnings
management by also studying discretionary accruals and income smoothness.
Other papers such as from LaGore et al. (2011) have examined if the amount of restatements within companies influenced the amount of CSR disclosures and found a relationship between those two but they did not investigate if a higher level of CSR within companies could reduce restatements. Chih et al. (2007) found that a focus on CSR lead to less earnings smoothing and less avoidance of earnings decreases and losses but increased earnings aggressiveness by studying four earnings measures for discretionary accruals. These studies show
7 that the effect of CSR on the quality of the financial information is not yet clear because the results conflict. Because the papers written on this topic are not that common and they are relatively new, this study investigates if the earnings quality differs among CSR and non-CSR firms. The research question is, therefore: “Does Corporate Cocial Responsibility influence the
earnings quality of a company?”
This study tries to answer this question by comparing three forms of restatements (restatements from SEC-investigations, accounting restatements and internal control weaknesses) as proxies for earnings quality among CSR and non-CSR firms. The sample of this study comprises U.S.
companies listed at the S&P 500 from 2003 to 2009.
The results of this research are that the CSR-score had significant influence on the internal control weaknesses of a firm and especially for SOX-302 errors but not for other forms of restatements. The CSR-firms also scored better on the corporate governance indicators. This indicates that CSR-firms tend to have better earnings quality than non-CSR firms because of their higher quality internal controls as those controls are associated with higher earnings quality (Ashbaugh et al., 2008). These findings are similar to the findings of Kim et al. (2012) suggesting that managers of CSR-firms provide more qualitative information to investors through the year report.
The contribution of this study lies in the fact that different measures of earnings quality are used to ascertain the effects of CSR on earnings quality. Previous papers such as Yip et al. (2011), Prior et al. (2008) and Chih et al. (2007) focused solely on the use of discretionary accruals and real time earnings management factors. Only Kim et al. (2012) have looked at the amount of AA-ers started by the SEC as a measure of earnings management. Through the use of three different proxies for earnings quality, this paper tries to give a clear answer on the effects of CSR on earnings quality. Also, the approach of Mattingly & Berman (2006), which involves separating CSR-strengths and concerns instead of adding them up as one score, is used to check the robustness of the results.
The next paragraphs consists of the literature review where the main concepts of this research and prior literature are discussed in greater detail and the hypotheses are given. The third paragraph contains the research design and methodology. The fourth chapter withholds the results of this research and this paper ends with the conclusion in the fifth chapter.
8
2. Literary review and hypotheses
This chapter starts defining CSR from a historical perspective which will help in explaining why the need for CSR nowadays, is seen as essential for companies to survive. When that has been discussed, the relationship of CSR and earnings quality is explained by the use of several
theories, concepts and prior research on the topic. This chapter ends with the development of the hypotheses of the research.
2.1 Corporate Social Responsibility
Corporate social responsibility has been the subject of many studies when companies started to embrace other obligations than the obligation to make profit for shareholders. Because of these studies, the definitions and point of views on CSR have also evolved throughout the years. Lee (2008) and Carroll (1999) describe the historical evolution of the literature on CSR and they both start with the first theoretical framework of CSR by Bowen (1953). Bowen (1953) was the first to develop a framework for CSR. He defined CSR as the obligations of businessmen to pursue those policies, make those decisions or follow those lines of actions which are desirable in terms of the objectives and values of society (p.6). Bowen argued that because companies have such a great impact on society, they need to assess the social consequences and impact of their actions. This statement was from a normative point of view. Bowen’s view was that companies had to embrace their social responsibilities because it was the right thing to do.
This philosophy was criticized and the need for CSR was controversial because
embracing social responsibilities, defining actions to meet social problems and developing social policies costs money and affects the bottom-line performance of companies. Friedman (1962), for example, proclaimed that the social responsibility of corporations is to make money for its
shareholders and therefore CSR is a subversive doctrine which could threaten free enterprise society. He was afraid that shareholder funds were misused by opportunistic managers under the name of CSR and that managers did not have the skills and expertise to deal with CSR.
Wallich & McGowan (1970) proposed a view in favor of CSR. They argued that CSR does not interfere with the interest of shareholders because shareholders hold diversified portfolio’s to
9 decrease risk and increase profit maximization. This means that they would be interested in profit maximization for all companies together, not a single company. Therefore companies need to strive for the social optimum because in the end, this would generate the most profit for
shareholders and companies in the long run. This approach to CSR is called the enlightened self-interest view and several papers have developed and evolved this theory on CSR.
An example of the self-enlightened view is the paper of Fitch (1976) who defined CSR as “The serious attempt of companies to solve social problems caused wholly or partly by the corporation ” (p.38). Fitch (1976) thought that companies should identify and assess social problems and decide which social problems should be dealt with first. This definition is quite vague so to specify what CSR can mean through a more detailed perspective, the definition of Caroll (1979) is presented.
Caroll (1979) thought that economic and social goals were not inherently conflicting with each other in contrast to what Friedman (1962) suggests. He argued that “the social responsibility of business encompasses the economic, legal, ethical and discretionary expectations that society has of organizations at a given point in time” (p.500). This meant that the company not only had to sell goods and services at a profit, as the economic perspective suggests, but companies also had to embrace other social values which is similar to the definitions of Bowen (1953) and Fitch (1976). However, the definition of Caroll (1979) is more specific and focuses on defining the social perspective into several perspectives which relate to the self-enlightened view, defined by Wallich & McGowan (1970).
The first perspective is the legal perspective which means that society expects companies to obey the law. The second view is the ethical perspective which means that society expects companies to show certain behavior which is conform to the ethical norms of society. The last perspective, the discretionary perspective, encompasses all other activities companies engage in, but which are not directly an expectation of society. Engaging in these activities can provide a strategic advantage to the organization because the company can distinguish itself compared to other companies. These activities can consist of voluntary activities, philanthropy and helping society in general.
Another view on CSR is given by Jones (1980) who defined CSR as “the notion that corporations have an obligation to constituent groups in society other than stockholders and beyond that prescribed by laws and union contract” (pp. 59-60). His contribution to defining CSR
10 lies in the fact that he saw CSR as a continuous process because the expectations of society of what defines socially accepted behavior, are constantly changing. Companies must therefore voluntarily adopt policies to help them meet the standards of socially accepted behavior and keep reviewing if these policies still meet the expectations of society. Therefore, companies need to assess if their corporate social performance is of sufficient quality to meet those expectations. Soial performance can be defined by the three-dimensional model of Caroll (1979). He defines corporate social performance through the use of three dimensions. He states that managers need to understand what CSR means to their company (responsibility), understand the issues where the company bears responsibility to its key stakeholders (issues) and management needs to specify how they respond to these issues (responsiveness).
Wartick and Cochran (1985) have developed the model of Caroll (1979) by arguing how companies can develop ways to meet the three dimensions of responsibility, social issues and responsiveness. Their framework states that companies need to adopt principles, processes and policies which are in accordance with the expectations of society. Principles can be used to meet the ethical responsibilities of the company, processes can be adopted to meet the responsiveness criteria and policies can be adopted to meet the social issues of the company.
Besides the self-enlightened view, Johnson (1971) outlines four other perspectives on CSR. The first view he presented, was the conventional wisdom perspective which defines a socially responsible firm as one whose managerial staff balances a multiplicity of interests. Instead of only striving for a larger profit, companies should take into account employees, suppliers, dealers, local communities and the nation (p.50). This can be seen as an early form of the stakeholder theory where companies operate within a socio-cultural system that outlines norms and business roles for responding to particular situations and ways of conducting business.
The second view regards social responsibility as a way for companies to add profits to their organization by carrying out social programs (p. 54). This view regards social responsibility as a way to maximize long-term profits. The third view views companies as utility maximizers and so companies do not only strive to increase their profits but have other goals as well (p. 59). Socially responsible managers do not only strive for their own goals but strive for the utility of other members of the enterprise and fellow citizens as well.
11 several goals in order of importance (p. 75). Targets are set for each goal and the company wants to score at least as well on those goals as other companies score on those goals. This view
suggests that companies who strongly focus on profit, act if social responsibility is important to them until their profit goals are attained. In reality, social goals are not important to them.
Johnson (1971) considers these four view to complement each other in assessing why companies adopt social responsibilities.
Later studies began to focus on specific aspects of CSR as the concept became widespread. This lead to CSR being split into several concepts but also to the development of new views on CSR involving stakeholder theory and strategic CSR (Lee, 2008). New concepts which were added were business ethics, public policies, corporate social responsiveness and stakeholder theory & management (Caroll, 1999).
The stakeholder theory of CSR means that companies have to identify the most important stakeholders whose cooperation is essential for the company to survive such as customers, suppliers, employees, the government and actions groups like the WNF (Freeman, 1984). The essential theme is survival for the company and these stakeholders all demand different things from the company. In order to survive, the company must meet these demands.
Clarkson (1995) argued that in order for managers to assess their social performance and so meet the demands of stakeholders, they need to distinguish between social and stakeholder issues. Social issues are public issues that are so serious that eventually legislation and regulation is necessary. Stakeholders issues are social issues for which no regulation is necessary. Managers should focus on social issues and leave stakeholder issues alone. By analyzing issues this way and by conducting this on an organizational, individual and institutional basis, managers can effectively assess the social performance of the company and meet demands of the public. A variant of the stakeholder theory, the instrumental stakeholder theory, suggests that CSR can be used to meet the demands of stakeholders by showing commitment to values which
stakeholders deem important. These demands can be managed by incorporating CSR (Lee 2008; Jones 1995). The stakeholder theory is an important theory because it provides a framework for managers to identify which demands the company needs to address. This theory also explains how CSR can increase the financial performance of a company by embracing the values demanded by powerful stakeholders. The theory does not differentiate between economic and
12 social goals of the company because both are needed to survive. This view that those goals
cannot be separated anymore began to dominate as CSR became rationalized (Lee, 2008). This happened because CSR became a practice to increase the competitive advantage of the company over others and to increase bottom-line performance instead of only a moral
responsibility (McWilliams et al., 2006). This change is also caused because the public started to demand for more social values within companies and became more aware of the effects of companies on the environment. This lead to corporations putting more pressure on behaving in a socially accepted manner so that they can perform better financially (Lee, 2008). Therefore management of a firm requires more attention today to various forms of performance, financial and social issues.
This attention for CSR lead to the broadening of the scope and the definition of CSR which forced researchers to specify the different concepts of CSR more clearly to understand the general concept (Lee, 2008, p. 62). New features of CSR were added such as corporate citizenship, embracing business ethics, strategic philanthropy, environmental responsibility, transparent accounting practices and affirmative action (Lee, 2008; Carroll, 1999). Corporate citizenship can be defined as the aim of companies to generate a higher standard of living and quality of life for the communities where they operate and live in, while maintaining profitability for stakeholders.
Business ethics can be defined as the expression of the company to commit to ethical values in their ways of conducting business through the use of ethical codes or social
responsibility charters. Strategic philanthropy can be seen as an activity of a firm that involves choosing how it will, voluntarily, allocate resources to charitable activities. This is done in order to reach marketing and business related objectives for which there are no clear social
expectations as to how the firm should perform (Ricks, 2002; Ricks & Williams, 2005).
Environmental responsibility is the practice of companies to show more care towards the environment that they operate in and to limit their impact on it. Transparent accounting practices are practices of showing shareholders more clearly how the company created value for them. This can involve, for example, the use of better measurement techniques or disclosing more
information (Jang, 2006, p. 168). Affirmative action programs are programs to promote the hiring and admission of minorities in society (Garrisson & Modigliani, 1994, p. 373). All these aspects
13 of CSR can be seen as programs to promote ethical and social values throughout the company and to outsiders to increase the competitive advantage of the company.
The last point view discussed in this literature is the concept of strategic CSR which can be seen as one of the leading views on CSR today. This concept is defined in a study of Porter & Kramer (2006). Strategic CSR defines CSR as “all the activities the company engages in while doing business”. Strategic CSR rejects the idea of a conceptual break between economic performance and social performance because both are essential to gain a competitive advantage. This means that CSR cannot only be seen as a moral responsibility but also as essential for survival.
Therefore, companies should not only try to limit their harmful effects on the value chain but they should also put together small number initiatives with large and distinctive social and business benefits (Porter & Kramer, 2006, p. 10). Porter & Kramer (2006) illustrate this with the example of the Toyota Prius. The Toyota Prius is an example of strategic CSR as Toyota
incorporates environmental friendliness and creates a competitive advantage through the development of this technologically advanced car. The authors suggest that these developments could lead to self-sustaining solutions for social problems such as global warming.
This chapter ends with a more practical example by looking at what CSR can withhold for a large company, namely Imtech, a technical service provider in the fields of electrical service, ICT and mechanical service. Their Sustainability report of 2013 states that to them CSR means “helping to meet the world’s growing economic, environmental and social needs in responsible ways. By focusing on business-related topics we can make a difference” (Imtech, Sustainability Report 2013, p.9).
Their CSR strategy covers four domains, market, environment employees and society. Through the market domain, Imtech wants to develop and offer new, innovative solutions to help customers develop sustainability goals. They do this by focusing on green products through their Greentech label and Codes of Sustainable Supply. The environment domain focuses on
measuring the CO2 emission of the whole company and reducing these negative effects. The
employees domain is about facilitating a safe work environment and supporting employee health. Through business ethics and health & safety standards, they aim to create an open and transparent business culture. The last domain is society which focuses on communicating with stakeholders.
14 This is important to ensure Imtech’s corporate citizenship, maintain their social license to operate and to help Imtech to solve social challenges.
This reports shows that several concepts which have been described before, come back into the social report of Imtech such as the focus on stakeholders, the concept of corporate citizenship, environmental responsibility and the social license to operate. The last concept is a key concept for the legitimacy theory which will be discussed in the next chapter.
2.2 Earnings quality
Earnings quality can be described as to what extent the financial report provides information about the financial performance of a firm, that is relevant to a specific decision made by a
specific decision-maker (Dechow et al., 2010, p. 344). Another definition, given by Francis et al. (2003), states that accounting earnings should reflect material changes in wealth. Therefore, earnings quality depends to what extend the actual material wealth changes of a company are reflected within the accounting earnings. Francis et al. (2008, p. 1) defines higher earnings
quality as more precise with respect to an underlying valuation relevant construct, that earnings is to describe.
These definitions show that earnings quality can be considered to be the true value created by the firm which should be reported through the year report. Earnings quality relies on two
components, this fundamental performance and the accounting system that measures this performance. Dechow et al. (2010) present three determinants of fundamental performance.
The first determinant is the cash flows generated during the current period, the second is the present value of cash flows that will be generated in the future as a result of actions taken in the current period. The third determinant is the present value of the change in the liquidation value of net assets as a result of actions taken in the current period (Dechow et al., 2010, p.347). Examples of these three determinants are, for the first determinant, direct cash received from a customer. The second determinant could be accounts receivable and the third the depreciation of a building or equipment. Those last two categories are the accruals and they represent an
acceleration of the recognition of future cash flows in earnings so that the timing of the
accounting income matches the timing of the inflow of economic benefits or drawbacks (Dechow and Dichev, 2002, p. 36).
15 The second component of earnings quality is the accounting system that measures and reports this fundamental performance through the annual year report. This system depends on which decision model the standard setters choose and what measures they use for the
performance of the company (Dechow et al., 2010). If the standard setters determine that only the earnings should be the reported performance, certain cash flow changes could be left out. The choice of measurement technique of earnings could also change the quality because there are different ways to measure earnings such as fair value changes of the company or the
sustainability of the cash flow changes.
There are two perspectives for the choice of measurement techniques within the
accounting system, the information perspective and the decision perspective. Francis et al. (2008) describe these perspectives as the purposes of financial reporting. Their first purpose, the decision usefulness perspective, states that high quality financial reporting should lead to more quality judgments and decisions. The second perspective, the information perspective states that shareholders can assess, through the year report, how well managers are doing their assigned tasks. This means that managers can assess their own performance and share/stakeholders can judge the manager on that performance. If the accounting system measures the created value of the company better, the year report and the earnings are more relevant to serve these purposes and the earnings quality is higher.
Three errors can arise when an accounting system tries to measure the fundamental performance which can causes differences in earnings quality among firms (Dechow et al., 2010). The first is that there are multiple decisions models within an accounting system and standard setters evaluate standards among the needs of users and so the information is not perfectly relevant to each individual shareholder. The second is that firms can choose between measurement
techniques which are already determined by the standard setters. This means that differences can arise because those techniques do not provide the perfect way to measure the fundamental performance. The third difference arises because measurement techniques are always subject to estimation and judgment error which can be due to unintentional errors or intentional bias. Unintentional errors are caused by the accountant and the manager who both need to value, measure, recognize and classify business transactions. Because of this judgment involved, errors can arise.
16 attempt of corporate officers to influence the short-term reported income (Sevin & Schroeder 2005). According to the agency theory, managers have an incentive to report higher earnings because managers are judged on their ability to create value for the shareholders because they are often extra rewarded on the basis of this value creation (Jensen and Meckling. 1976). Therefore, if a manager exercises his discretion upon the accounting numbers and inflates the numbers, he is more likely to reach a bonus set out by the shareholders of the company.
This way the earnings quality of the company decreases because the reported accounting numbers relate less to the actually created value of the company. There are in general four reasons for practicing earnings management (Beatty and Weber 2006; Sevin & Schroeder, 2005; Healy & Wahlen, 1999). The first reason is to meet shareholders expectations because if the manager cannot live up to the expectations of shareholders, the share price might drop and the manager can be judged on that. The second reason can be to prevent the violation of debt covenants as the costs of breaching those contracts can be quite costly to the firm and the manager (Beatty & Weber, 2006, p. 264). The third reason can be to increase the manager’s own compensation because most compensation of CEOs depend on equity and by inflating the numbers, share price can increase. The fourth can be to prevent governmental regulation.
Five techniques are most common for practicing earnings management as identified by Sevin & Schroeder (2005, p. 49). The first is “taking a bath” which occurs when a company takes a one-time large charge against income to reduce future expenses on assets. The second is “creative acquisition accounting” which is the practice of avoiding future expenses by one-time charges for in-house research and development. The third is “cookie jar reserves” which involves
overcharging costs like sales expenses or warranty costs so that the overcharged part is
supposedly to offset an increase in claims even though the normal size of the warranty costs and sales expenses would be enough to settle the expected future claims. The fourth is “abusing the materiality concept” which involves recording small errors and mistakes because their impact is not of enough importance for the financial statements. The last one is “improper revenue
recognition” which involves recording revenues before the point that they are properly earned. These earnings management techniques all dilute the quality of the financial statements but only if their impact is material, they will lead to restatements. Restatements mostly happen
17 when the statements of the company have been pushing the boundaries of fair accounting so they can be considered somewhat extreme cases (Richard et al., 2002, p.2)
2.3 CSR and earnings quality
Several research papers have been written on the investigation if CSR could have an impact on earnings quality. This literature can be can be roughly divided into two streams; literature
predicting an increase in earnings quality if a company focuses on CSR and literature predicting a decrease in earnings quality. This chapter starts with discussing the literature which suggests that CSR-firms could have higher earnings quality compared to non-CSR firms.
Theories predicting higher earnings quality for CSR-firms base their predictions on the fact that if managers are truly committed to the ethical values they proclaim, they will also act more
ethically when they produce the numbers in the year report. CSR can be seen as a moral responsibility to stick to those values (Porter and Kramer, 2006).
This can be considered the same as the ethical responsibility of companies which Caroll (1979) describes. His view on a company engaging in CSR was that the company should strive to make a profit, be a good corporate citizen, behave ethically and obey the law. He defines acting ethically as acting in accordance with the values society deems important besides legal
requirements. These values are considered ethical norms and an ethical norm stakeholders find to be important is that the company is honest and trustworthy (Lee, 2008). Business ethics is an important part of CSR and diluting the quality of earnings through practices such as earnings management is generally considered to be unethical (Kaplan, 2001). As earnings management is a mild form of misleading the fundamental performance of a firm, ethical managers could be hesitant to engage in this sort of behavior (Kim et al, 2012).
Lee (2008) suggests that part of CSR is the manager’s responsibility to be more transparent in the accounting methods used to establish the year report. Therefore, the earnings quality should be higher for firms with a higher commitment to CSR values. Jones (1995) hypothesizes that because a commitment to ethical behavior can be beneficial to the firm, managers can be incentivized to act according to ethical behavior standards. He states that moral firms embrace values which will not fit the value images of opportunistic people. This means that opportunistic managers will leave the firm or be fired out of fear for reputation damage or because their moral
18 values are questioned. As opportunistic behavior is more closely related to earnings management, moral managers might be more likely to refrain from these practices.
Another reason to observe a higher quality of earnings for CSR firms is that consequently
managing earnings upwards can lead to the destruction of the core values of the company (Jensen et al., 2004). This situation arises when earnings management leads to the overvaluation of equity, because the equity in the books is higher than it really is due to the higher reported earnings. This means that the expectations of investors and analysts will be higher and they will require higher a returns on equity from the company to meet these expectations. This increases agency costs and eventually, the performance of the company will be unable to keep up with expectations and earnings management practices will be disclosed.
The disclosure of these practices can have severe negative consequences for the
reputation of the company and these consequences will be more severe for CSR firms because of their proclaimed ethical and social values (Linthicum et al., 2010). Because of this extra negative consequences for CSR-firms, the managers of those firms might be more likely to refrain from earnings management. Therefore for the aforementioned reasons, the first hypothesis of this research is:
H1: A company with a strong focus on CSR is more likely to have a strong quality of earnings.
The opposite can be said just as well, several theories provide evidence on why CSR might be a disguise for the lower quality of earnings of companies. Those theories state that CSR could be used to make up for the low quality of the financial statements. Studies from Jensen and Meckling (1976), Carroll (1979) and Friedman (1970) hypothesize that the manager can use CSR for this own personal good. The motives, suggested in these studies, are that managers adopt CSR to help their careers or to pursue own motives instead of the motives of the shareholders.
A theoretical framework for these reasons can be provided through the legitimacy theory. The legitimacy theory states that companies have a social contract with society because society provides organizations with a social license to operate if they act according to the bounds and norms society sets (Brown and Deegan, 1998). In order to be “legitimate” society needs to have the generalized perception or assumption that the actions of an entity are desirable, proper or appropriate”(Suchman, 1995, p. 573). If the operations of organizations can no longer be
19 justified, society can revoke this license by refusing to buy the products of the organization, refusing to supply the organization with the goods the organization needs for survival or the government can prohibit the operations of the organization. Therefore, organizations have an incentive to operate within the bounds and norms of society or to at least operate within the bounds and norms of society in the eyes of society (Brown and Deegan, 1998). Any departures from the bounds and norms of society, will only be accepted if they are considered to be unique and one-of-a-kind cases (Perrow, 1981). If those departures happen on a frequent basis, the license to operate of the organization can be withdrawn by society.
Organizations can seek legitimacy for many reasons but Suchman (1995, p. 574) distinguishes those reasons for legitimacy between seeking continuity versus credibility and seeking active versus passive support. Actions to seek legitimacy through continuity aim at the fact that organizations that appear more desirable and appropriate are more likely to receive resources from different groups in society. This aim focusses on the fact that those groups will keep coming back more often for the services of the company. Therefore, the organization is more likely to continue its operations and will be more stable.
Legitimacy gained from credibility is when the customer not only acts favorably towards the organization, but understands the organization. A legitimate organization in this way is more meaningful, understandable, predictable and more trustworthy to different actors in society. This process arises because shared understandings lead to the rationalization of the behavior (Pfeffer, 1981). These two forms of legitimacy can be seen as creating persistence versus meaning. The other way companies can seek legitimacy is to whom the organization seeks active support and to which groups passive support is needed to continue its operations Suchman (1995, p. 575) . Active support can be, for example, essential to survive in a highly competitive market because customers can be hard to retain. However, some groups in society could be left alone because the company does not need their active support for survival. An example can be customers which are not profitable enough for the company to spend resources on.
Legitimacy can be split into three different forms, namely pragmatic, moral and cognitive
legitimacy (Suchman, 1995, O’Dwyer et al., 2011). These three types differ on what behavior the different interest groups or constituents demand from the company and what behavior the
20 company displays to the constituents it seeks legitimacy from. Pragmatic legitimacy is when the interest group deem the company legitimate when it provides value to them. This legitimacy can arise through three ways. The first is because of particular policies of the company which the interest groups support. The second is because the company appeals to the larger interests of the interest group which is often done through including and involving constituents in the policy setting or performance standard setting processes of the company. The third way is because the company convinces the constituents that the policies and practices of the company are in the interest of the constituents and can be trusted. This last way can be seen as a widespread or naïve belief that company is legitimate (O’Dwyer et al., 2011).
The second type of legitimacy is moral legitimacy, which can be seen as a positive and normative evaluation of the company and its activities (Suchman, 1995). Interest groups deem the activities of the company not only as valuable to them, but they consider those activities the right thing to do. The actions of organizations are often perceived as the right thing to do when they effectively promote societal welfare. This social welfare is determined by the audience’s socially constructed value system (O’Dywer et al., 2011, p. 36).
The audience evaluates the moral aspect of the company through the evaluations of what outcomes the company achieves, what procedures the company follows, which leaders and staff is employed within the company and what social constructs are apparent within society which legitimize the tasks of the company. An example of such a social construct can be a, perhaps state-owned, company with a monopoly on telecom services because the government wants to have a own telecom network in times of war against other countries. The company is then legitimatized by the government to have the monopoly on telecom services.
The last form of legitimacy is cognitive legitimacy which is achieved when constituents
automatically assume that the activities and objectives of an organization are proper and desirable and therefore, legitimate (Suchman, 1995). This is very hard to achieve for a private organization but this could relate to the government of a stable country. The government is accepted by
inhabitants to be the only organization to possess the right to use violence to discipline all inhabitants of the country to obey the law. Because of this legitimacy, the government may use violence or prison sentences to ensure that all inhabitants obey the law.
21 the organization needs to make clear why it is essential that the organization exists and why it is the only organization that can achieve the objectives set out. This depends on the fact that if the company would seize to exist, the activities of the company would collapse and this would provide a perceived loss of value to constituents (Suchman, 1995, p. 582). The second feature of cognitive legitimacy is taken-for-grantedness which is such a subtle form of legitimacy that a removal of the social structure would be unthinkable (Zucker, 1983). Therefore, the constituents do not think about the fact that other organizations can also perform the activities of the
organization which is taken for granted. This way, the organization that is taken for granted, has no competition from other organization because the interest groups do not take other
organizations into consideration when assessing which organization is best for performing certain tasks or activities.
These forms of legitimacy can be gained or kept by using various strategies which are identified by the framework of Suchman (1995) and Lindblom (1994). Lindblom (1994) identifies four strategies for creating and or maintaining legitimacy which consist of:
Seek and inform society about the changing operations and activities of the organization. Change the perception of the organization but not actually change the actual behavior of
the organization.
Deflect the attention of the operations of the organization by focusing on other issues and adopting symbols of legitimacy. A symbol of legitimacy is a symbol which the
organization adopts to show that they comply with the expectations of society and to show their legitimacy.
Change the expectations of society so that these expectations fit the operations of the company.
The strategies of Suchman (1995) are similar to the strategies identified by Lindblom (1994), but Suchman focuses on gaining, maintaining and repairing legitimacy and relates his strategies to the various forms of legitimacy. He identifies three strategies for gaining legitimacy which can differ for what type of legitimacy the company seeks. The first strategy Suchman (1995) identifies, is:
22 - Conform to the expectations of constituents in the current organizational environment. This strategy consists of complying to the various expectations of the interest groups of the organization. To achieve pragmatic legitimacy, the company must meet the substantive needs of the constituents or offer them access to the process of decision making within the company so that the instrumental demands of those groups are met. When these needs are met, or the constituents think that these needs are met, the company is deemed to be legitimate, in a
pragmatic way. To achieve moral legitimacy, the company must show that they comply with the principle ideas of the constituents such as that an organizations should be a good corporate citizen. To achieve cognitive legitimacy, the company must meet the standards or social model of the constituents (Suchman, 1995; O’Dwyer et al. 2011). An example can be that a mining
company needs to have a permit approved by the local government to start mining in a certain area. The permit is a symbol of complying to the values which the government and society deem important to show that the organization is legitimate to perform these activities.
The second strategy of Suchman (1995) is:
- Select among multiple environments which are more likely to support the activities of the organization.
The strategies are similar, for pragmatic legitimacy the company must select new audiences whose instrumental values relate to the organizational activities. For moral legitimacy, the company must find new constituents whose moral values appeal to the organization and for cognitive legitimacy, the company must proclaim a new set of values which are already accepted in a related area (O’Dwyer et al., 2011). The third strategy is:
- Manipulate the environment by creating new legitimate beliefs and new audiences. This strategy is not only about manipulating demands of society by showing commitment to values society deems important, but also about creating new explanations for social reality to achieve an alternative reality which supports the company (Suchman, 1995). This means that the organization tries to provide new reasoning for the need for their activities to create an alternative social reality in which the activities of the organization are perceived to be legitimate.
To achieve pragmatic legitimacy through this manipulation, the organization must use strategic communication to appeal to, and persuade new audiences to value the organization and
23 their activities. This is done to make the audiences believe that they can influence the
organization while in reality, they cannot. For moral legitimacy, the company needs to focus on showing a record of technical successes and demonstrating how it achieved this socially desired performance. This is done through techniques and procedures which provide a basis for the four categories of legitimacy. This works even better if this is done through a collective of similar organizations.
To achieve cognitive legitimacy through manipulation, organizations need to collectively, adopt similar practices so that the activities and practices of the organizations within this group, will become taken for granted. Companies can also try to achieve this through collectively informing the public about the supposed activities of the organization. This is done in order to raise a manipulated understandability of the activities the organization and can be achieved through lobbying and advertising.
These strategies range from conforming to the demands of constituents to active manipulation which shows that CSR activities could be adopted for multiple reasons and to achieve multiple forms of legitimacy. Changing the perception of society about the operations of the organization, can be done, for example, by adopting CSR values as a symbol of legitimacy (Brown and
Deegan, 1998). Because CSR reports promote the ethical values of the company and the CEO, it might be a form of deflecting attention from unethical actions of the company such as managing earnings. As managing earnings is in general considered to be an unethical practice, it could threaten the legitimacy of the organization (Kaplan, 2001).
If we relate CSR as a legitimizing strategy for unethical practices to the framework of Suchman (1995), we can see that in order for this strategy to work, an organizations needs to engage in strategic communication, promotion of social values, showing a track record of technical successes and advertising about the activities of the organization, in order to achieve pragmatic, moral or cognitive legitimacy.
These activities all are generally found in a sustainability report such as the Imtech report, mentioned earlier. The values promoted in this report were innovative solutions for sustainable goals, business ethics, health and safety standards, reduction of CO2, business ethics and
corporate citizenship. Imtech strategically communicates these values to their shareholders and they have numerous track records of technical successes, mentioned in this report.
24 This could be a way of showing that the company conforms to the expectations of society but it could also be explained as a way of manipulating the environment of the organization. This suggest that the earnings quality could be lower for CSR-companies because managers put a stronger focus on CSR to legitimize the lower earnings quality of the organization (Prior et al. 2008). This relationship is hypothesized by Prior et al. (2008, p.162) who states that CSR can be used to gain favorable coverage from the media, legitimize themselves to the community, gain favorable regulation and avoid scrutiny from employees and investors. Managers could this way be incentivized to focus on CSR so that they can avoid demands of shareholders to correct earnings management practices.
A method companies can use to legitimize their actions, is through the disclosure of information which justifies their continued existence (Lehman 1983). The annual report can be seen as the key form of disclosure of companies as it displays the proclaimed policies of a company. These polices represent a response to the actual or perceived demands of particular interest groups for more information from the company (Deegan and Gordon, 1996; Lehman, 1983). A study of Guthrie & Parker (1989), for example, illustrates this as they found that environmental disclosures of mining, oil and steel industries increased when they were under increased pressure from environmental action groups. These action groups were criticizing the pollution caused by the companies and so, these companies were trying to counter the criticism of actions groups in order to remain legitimate in the eyes of other interest groups of society.
There is also criticism on the legitimacy theory as it might not be applicable to all industries such as fitness, day care and grass-root politics (political systems which are just developing, often with involvement of citizens). This inability to use this theory arises from the fact that the
organizations in these industries cannot properly define the socially desired outcomes. This means that the proper actions, the organizations should follow, cannot be determined because the causality with the outcomes is unclear or there is no proper behavior to adopt.
The second theory which could prove reasoning for lower earnings quality is the stakeholder theory. This theory is quite similar to the legitimacy theory as it also states that the organization needs to please society so that it can keep operating. The difference is that stakeholder theory states that only the demands from the stakeholders, who are powerful enough to have an impact
25 on the resources of the organization, need to satisfied (Donaldson and Peterson, 1995). A
stakeholder can differ in importance to the firm and can differ in the power they have over the manager. Satisfying the demands of powerful stakeholders is essential for organizational performance, according to this theory (Hill and Jones, 1992).
The stakeholder theory can be divided into two parts, the instrumental stakeholder theory and the normative stakeholder theory (Donaldson and Peterson 1995). The instrumental
stakeholder theory states that the demands of the stakeholder with the power to influence the achievement of the goals of the organization need to be satisfied. Normative stakeholder theory states that the demands of all stakeholders should be met. If CSR values would be adopted by a manager in order to manage the demands of stakeholders then, according to the instrumental stakeholder theory, a manager should only focus on the values demanded by important
stakeholders and ignore other stakeholders. This way other demands such as for better earnings quality could be suppressed.
The institutional theory is a similar theory which studies how routines and norms are adopted within organizations (Dimaggio and Powell, 1983). Both the institutional theory and the legitimacy theory suggests that companies might have completely different routines and norms than the ones they express to the outside world. This could mean that the promoted ethical values through CSR are a camouflage for the low earnings quality which forms the second hypothesis of this research, namely:
H2: A company with a strong focus on CSR is more likely to low quality of earnings.
The theoretical basis of this research has been discussed in this chapter, the related literature is discussed in the next paragraph.
2.4 Related literature
Research supporting higher quality earnings for firms with a stronger focus on CSR, can be found in the papers of Kim et al. (2012), Chih et al. (2007), Andersen et al. (2012), Scholtens & Kang (2012) and Yip et al. (2011). However, other papers suggesting lower quality earnings for firms with a strong focus on CSR, are also common such as papers from Prior et al. (2008), Setyorini & Ishak (2012) and papers finding no statistical relationship can be found such as Sun et al. (2010).
26 Kim et al. (2012) found, by looking at several proxies of earnings management such as
discretionary accruals, real earnings management, GAAP-violations through AA-ers and using CSR performance scores, that CSR firms in the U.S. are less likely to engage in aggressive earnings management and are less likely to be sanctioned by the SEC for a violation of GAAP. They, however, looked only at one category of restatements, the SEC violations and did not investigate if internal control deficiencies and actual restatements are influenced by a focus on CSR.
Chih et al. (2008) found that companies listed at the FTSE-Global Good index, had less earnings smoothing, less avoidance of earnings decreases and losses, but increased earnings aggressiveness. They found this through studying four earnings measures for discretionary accruals for 1653 companies in 46 countries through the FTSE and the FTS4good-index.
Anderson et al. (2012) found a positive relationship between socially responsible firms and transparent accounting, as well as less earning smoothing but increased earnings
aggressiveness for CSR firms, but found differences for these results among different industries. Their conclusion was that the effect of CSR on earnings quality is related to corporate
governance.
Scholtens & Kang studied CSR and earnings management within ten different Asian countries and found similar results as the previous two studies. These results are that earnings smoothing is reduced and earnings aggressiveness increases which they relate to the different legal enforcement systems of Asian countries. Yip et al. (2011) found by studying CSR disclosures in the oil & gas industry, a negative relationship between the amount of CSR disclosures and discretionary accruals of companies. This implies that CSR firms have a higher quality of earnings. However, they found a positive relationship for the food industry. This shows that the relationship of CSR and earnings quality can differ among different industries. They relate this result to political considerations as the oil & gas industry is under much more political and ethical scrutiny than the food industry.
In contrast to previous studies, Prior et al. (2008) found that CSR increases earning smoothing. They found this result by studying the discretionary accruals of 593 multinational companies from 2002-2004 and using archival data on the Sustainable Research Development Index from 2002 – 2004. They find that income smoothing increases with CSR. They control for financial performance to show that managers do this mostly to increase short-term performance which they
27 relate to the attempt of managers to mitigate stakeholder activism by showing good behavior through CSR.
Setyorini and Ishak (2012) found, that if the bonus of managers is based on accounting numbers, they will disclose CSR activities in order to increase the performance of the company. Managers also were more likely to manage earnings when their CSR disclosures were higher which the authors relate to the reduction of political costs in India. This suggestions rests on the assumption that managers do this to prevent unfavorable judgment from the government or to reduce the unfavorable effects of a new ruling. They found this by studying companies listed at Indonesian stock exchange and by using data from the Clarkson’s Environmental Index (2007) and Sutanto Para’ social index.
Sun et al. (2011) use performance-matched discretionary accruals to study if CSR disclosures and earnings management are related in the U.K.. They use environmental reporting disclosures but find no significant relation between CED disclosures and earnings management. They do find that corporate governance is related to earnings management. This chapter
discussed the relevant literature on earnings quality and CSR, the next paragraph presents the research design to test these hypotheses and the results of this research method.
28
3. Research Design & Methodology
This chapter explains what methods and what sample is chosen to test if CSR has an impact on earnings quality. Through the use of a regression model, the hypotheses will be tested. The first section explains the sample selection and what measures are taken to define earnings quality. The second section explains which regression model is used to test the hypotheses. The last section describes how the sample was created.
3.1 Sample and variable selection
The sample for this research is based on archival data and is chosen in such a way that the effect of CSR on earnings quality is not caused by other factors. Three forms of restatements are chosen to proxy for earnings quality. Earnings quality has been defined earlier in this research as to what extent the financial report provides information about the financial performance of a firm, that is relevant to a specific decision made by a specific decision-maker (Dechow et al., 2010, p. 344).
Restatements can be chosen as proxy for earnings quality because restatements are external investigations which represent earnings misstatements (Dechow et al., 2010). After a material misstatement has been found, a company issues a statement of non-reliance for the restated year report to investors and the company may have to amend the inaccurate numbers. The general name for all types of restatements is therefore non-reliance restatements. The three types of restatements are represented in this study into three categories, the amount of AAERs by the SEC, the restatements of accounting numbers and the amount of Sox reports on internal control deficiencies within the companies.
Accounting restatements are restatements made by the company itself, restatements from SEC-investigations happen when the U.S. Securities and Exchange Commity accuses companies of overstated or misstated earnings (Dechow et al., 2010). Internal control deficiencies can be divided into SOX-302 and 404 errors. SOX-302 consists manager’s assessments of the
effectiveness of the internal control system of the company (Dechow et al., 2010, p. 375). This system consist of the necessary controls to guarantee the accurateness of the financial numbers. SOX-404 includes the manager’s assessment of the internal controls and the accountant’s assessment of the correctness of this statement by the manager. SOX-302 and 404 errors are when this internal control system is deemed to be ineffective by the manager or by the
29 accountant. This measure is chosen as proxy for earnings quality because if the internal control system is not effective, the correctness of the financial numbers cannot be assessed reliably. Companies experience an internal control deficiencies if they have a SOX-302 or SOX-404 error or both.
All three measures of restatements are used in the proposed research so that the effect of CSR on restatements can be properly aligned because there are benefits and disadvantages with each method. AAERs by the SEC are cases when the SEC accuses companies of overstating or misstating earnings and so have a low type-1 error because this way a lower quality of earnings can be directly measured (Dechow et al., 2010). However, the SEC needs to be sure that the manager has managed earnings and so the amount of cases is low because the SEC does not prosecute on ambiguous cases.
Accounting restatements and internal control deficiencies are weaker indications of earnings management but are more common and provide a measure of low quality of earnings (Dechow et al., 2010). The disadvantage with internal control deficiencies is that they are not mandatory disclosed so the disclosure might be subjected to bias. To control for these issues, all three types of restatements are used to ensure adequate measurement of differences in earnings quality.
The data for these three types of restatements are found through the use of the Audit Analytics database which provides data on non-reliance restatements, SOX-302 and SOX-404 errors. A timeframe from 2003 to 2009 is used to study the SOX-302 errors, accounting
restatements and restatements from SEC-investigations. Data in the AuditAnalytics database for SOX-404 errors only dates back to 2004 so a timeframe from 2004 to 2009 is used for this measure. 2009 is chosen as the latest year to get data from because of data requirements for the CSR-score.
The sample should have similar accounting standards to exclude the effects of different standards causing different restatements and that why the U.S. is chosen because they have one similar accounting standard, U.S. GAAP. Another reason to choose the U.S. is because it is one of the largest countries with one accounting standard so, the most data can be gathered. To meet the requirements of data on all three forms of restatements, I conduct my research on the companies
30 listed on the S&P 500 index. The ID’s for the companies used in this sample are obtained through the S&P Capital IQ Compustat North America.
By using the KLD database, CSR can be measured. The KLD-database can be seen as the ‘de facto research standard at the moment’ (Waddock, 2003, p. 369) and is used to measure CSR scores in various research papers such as Waddock and Graves (1997), Turban and Greening (1997) and has been tested by, for example, Mattingly and Berman (2006) for its usefulness. I use this database to measure the score on certain CSR strengths and concerns of companies in five different categories which are based on measuring CSR as stakeholder responsiveness (Berman et al. 1999; Mattingly and Berman, 2006). The KLD database evaluates companies on their
responsiveness towards stakeholders based on five categories: employees, community, diversity, environment and product quality and safety (Mattingly and Berman, 2006). Those categories are based on five stakeholder groups, employees, the community, diversity related stakeholders such as lesbian or feminist action groups, environmental action groups or funds like the WNF and the last measure, product quality and safety, is a measure for consumers as stakeholder group.
The scores on these categories are divided into companies with more strengths than weakness and companies with more weaknesses than strengths. The relevant variables for the CSR-scores are based on the variables used by Mattingly and Berman (2006) of which a list is provided in appendix A on pages 54 & 55.The chosen variables can easily be linked to strengths and weaknesses which is key to this research. However, corporate governance factors are
excluded as the construct of corporate governance and earnings management is a different relationship. The construct of this research is to measure the relationship of CSR and earnings management. Corporate governance and earnings management has, for example, already been examined in studies such as Ashbaugh et al. (2008), who looked at internal control deficiencies and the quality of earnings, and Cornett et al. (2008) who looked at earnings management and several corporate governance factors. Both studies conclude that firms with lower quality internal controls and corporate governance structures, have lower quality accruals which is a lead
indicator of earnings management.
This approach of excluding corporate governance factors matches the approach of Kim et al. (2012) who exclude corporate governance factors for the same reason. The results of the regression are checked by separating the scores on stakeholder responsiveness into one score for
31 strengths and one score for weaknesses. This way, I follow the approach of Mattingly and
Berman (2006) and Berman et al. (1999) who measure CSR performance as high or low stakeholder responsiveness instead of using one score for CSR which could obscure results.
The other financial data needed for the regression model presented in the next chapter is from the Compustat database which provides, for example, balance sheet and income statement data.
3.2 Model selection
The first test of this research consists of comparing if the three categories of restatements are more common with firms with higher or lower CSR ratings. This is done by comparing the firms within the three samples of restatements and dividing these samples into one group with more CSR-strengths than concerns and one group with more CSR-concerns than strengths. Through the use of a t-test, which compare the average amount of restatements among the two groups, the hypotheses are tested.
The second test consists of investigating if restatements are more or less common for firms with a focus on CSR, by using a regression model based on the models used by Prior et al. (2008), Sun et al. (2010) and Kim et al. (2012). This model can be used to test if CSR firms are more or less likely to experience restatements as proxy for earnings quality. By testing if the CSR score is positively or negatively significant and has influence on the independent variable restatement, the two hypotheses are tested. If this is the case, we can say that the company’s scores on CSR practices impacts their earnings quality.
The dependent variable for this regression is the form of restatement used and a
regression is done for all three types of restatements. The other variables, besides the CSR score, are independent control variables. These variables are used in comparable regression formula’s for papers which investigate such as Kim et al. (2012), Sun et al. (2010), Prior et al. (2008) and Chih et al. (2008) in order to reduce the interference of other variables with the relationship of restatements and CSR.
PR (restatements = 1) = 𝛼0 + 𝛼1 𝐶𝑆𝑅 𝑠𝑐𝑜𝑟𝑒𝑡+ 𝛼2𝑆𝑖𝑧𝑒𝑡−1+ 𝛼3𝑀𝑇𝐵𝑡−1+ 𝛼4𝑅𝑂𝐴𝑡−1+ 𝛼5𝐿𝐸𝑉𝑡−1+ 𝛼6𝑅𝐷_𝐼𝑁𝑇𝑡+ 𝑎7𝐺𝑜𝑣𝑒𝑟𝑛𝑎𝑛𝑐𝑒𝑡+ 𝛼8𝐵𝐼𝐺4 𝑡+ 𝑎9𝐴𝑢𝑑𝑖𝑡𝑜𝑟𝐶ℎ𝑎𝑛𝑔𝑒 + 𝜀𝑡 These variables stand for:
