Introduction

A growing body of management research has studied the relationships among ownership structure, strategic decision making, and firm performance. Previous research has found that ownership has an influence on organizational decision making, motivation, and power (Hart and Moore 1990; Finkelstein 1992). For instance, ownership structure affects important firm-level decisions such as R&D spending (Baysinger et al. 1991), innovation (Kochhar and David 1996), capital structure (Chaganti and Damanpour 1991), entrepreneurship (Zahra 1996), and diversification (Eisenmann 2002). In particular, the key owners (e.g., institutional owners, managerial owners, etc.) can go beyond the decision to invest or not to invest by proposing and voting on firm’s strategic decisions through multiple channels. Thus, given that corporate social actions can be seen as a form of “investment” (McWilliam and Siegel 2001), it is not surprising that the key owners are likely to be involved in the firm’s strategic decisions about social investments.

Corporate social responsibility (CSR) has received a significant amount of attention from both academic researchers and business practitioners. CSR is defined as corporate integrated responsibilities encompassing the economic, legal, ethical, and discretionary (or philanthropic) expectations that the society has of organizations (Carroll 1979). For over a decade, several economic and noneconomic determinants of CSR have been examined (e.g. Aguilera et al. 2007). Previous studies have found that external factors such as industry (Hackston and Milne 1996), customers (Vogel 2005), activist groups (den Hond and de Bakker 2007), regulation/law (Dawkins and Lewis 2003), and communities (Boehm 2005) affect organizations’ decisions about CSR participation.

At the same time, other studies have emphasized the role of organizational factors such as slack resources (Waddock and Graves 1997), firm size (Fombrun and Shanley 1990; Stanwick and Stanwick 1998), board structure (Johnson and Greenings 1999), top managers’ moral principles (Davis et al. 1997), and so forth. However, relatively few researchers (e.g. Barnea and Rubin 2010; Johnson and Greening 1999) have assessed the extent to which the ownership structure explains the firm’s social performance. In particular, since different owners may have different objectives and decision-making horizons (e.g., Hoskisson et al. 2002), it is worthwhile to study the relationships between the different types of owners and the firm’s social performance. In this article, we break down owners into three separate categories: institutional (pension funds, banks, insurance companies, and securities firms), managerial (top management team and outside directors), and foreign ownerships. We use this classification of owners because (1) institutional investors are traditionally large and may hold a substantial amount of a firm’s shares; (2) managers and directors are best-informed about the firm’s situation and often have the most significant influence on the firm’s strategy and investments; (3) foreign investors are likely to be distinct from domestic investors in their preferences, time horizons, and the extent of the information asymmetry problem. Given these differences, we expect different owners to have different preferences regarding the firms’ social investments.

Using a sample of large, public Korean firms, we found that ownership by banks, pension funds, and foreign investors has a significant positive relationship with the firm’s CSR ratings. Ownership by the firm’s managers, however, has a significant negative relationship with the firm’s CSR ratings while the effect of outside director ownership is not significant. Our results suggest that Korean managers, who are often associated with the Chaebol family (e.g. Baek et al. 2004; Chang 2003; Kim 2007), may use their stock ownership to obtain personal benefits at the expense of the other stakeholders. Banks, pension funds, and foreign investors may be forced to be long-term oriented because they cannot easily sell their shares without greatly affecting the stock price. In addition, foreign investors may be more inclined to pay attention to social issues because of their familiarity with these issues and greater emphasis on CSR in their home countries, as well as because of the role of CSR engagement as an important signaling mechanism that may reduce information asymmetry.

Theory and Hypotheses

In this article, we hypothesize that CSR participation is a result of decisions made by the corporate managers under pressure from the shareholders. We start our theory building with an overview of how an ownership structure may affect corporate decisions, including decisions to participate in social issues.

There are various mechanisms through which owners can affect corporate decision making. According to agency theory (Eisenhardt 1989; Jensen and Meckling 1976), managers are the firm’s owners’ agents that do not bear all the consequences of their own decisions. One of the problems with separation of ownership and control (Fama and Jensen 1983) is the suboptimal decisions that the managers might make in order to serve their own interests. If the owners of the corporation are dispersed and none of them own a significant percentage of the stock, none of the owners will have significant incentives to spend their time monitoring and evaluating managerial decisions. Even if some minor shareholders were willing to monitor managerial decisions, they would not have any means of affecting corporate decision making because they would not have the voting power on the board.

This situation changes when some shareholders own significant percentage of the stock. Large shareholders will usually have the power to affect corporate decisions via appointing directors on the board (Boyd 1994) and shareholder activism (Admati et al. 1994; Lee and Lounsbury 2011; Smith 1996). Thus, we would expect to see more active monitoring and intervention of shareholders into corporate decision making when some large shareholders own significant amounts of equity. Existing empirical research supports the argument that ownership concentration affects corporate decision making (e.g. Baysinger et al. 1991; Kochhar and David 1996).

An important question is how ownership structure affects the corporate decision making. By ownership structure, we understand the absolute and relative shares of the stock owned by the specific owners such as pension funds, banks, and investment firms, etc. Previous research has found that different types of owners have divergent preferences regarding various corporate decisions and investments. For example, Zahra (1996) found that executive ownership and long-term institutional ownership are positively associated with corporate entrepreneurship, whereas short-term institutional ownership has a negative effect on it; furthermore, outside director ownership is found to neutralize such a negative impact. In a similar vein, Chaganti and Damanpour (1991) found a significant negative relationship between outside institutional ownership and the firm’s capital structure (i.e., long-term debt-to-capital ratio), yet family and inside institutional owners moderate the relationship; in addition, executive ownership strengthens the relationship between outside institutional ownerships and financial performance including ROA, ROE, and P-E ratio. Importantly, Hoskisson et al. (2002) focused on how corporate innovation strategies are actually made by different institutional owners, examining the mediating effects of the roles of boards of directors. They documented that the key stock owners influence corporate strategic decisions directly and indirectly, arguing that directors in a corporate board are influenced by the owners and, in turn, themselves influence corporate innovation strategies.

Based on the existing evidence, we view CSR as a form of investment and hypothesize that varying shares owned by specific types of investors will have a differential effect on the corporate decisions regarding CSR. First, we view social investments as any other kinds of investments that are supposed to yield benefits for the shareholders (e.g. Cochran and Wood 1984). In particular, we assume that social investments lead to increased chances of long-term survival of the company, which will be positively valued by the stock market. There is evidence that socially responsible business is positively associated with the performance of firms, especially in the long run (Orlitzky et al. 2003), even though this relationship is not always reliable (Margolis and Walsh 2003). Our argument is built on the following premise. Socially irresponsible companies are subject to legal sanction and punishment from the government and other powerful stakeholders (Agle et al. 1999). Such actions against the company hurt its chances of survival in the long run. Therefore, it is optimal for socially irresponsible companies to improve their social performance by making charitable contributions, complying with regulatory requirements and being proactive about environmental issues, improving relations with local communities, etc.

Second, we argue that different shareholders may have different time horizons and thereby different preferences regarding corporate social investments. We assume that long-term shareholders are more likely to support social investments than short-term shareholders. This assumption is based on the following reasoning. CSR investments are most likely to pay off in the long run (Falck and Heblich 2007). In the short run, they might actually be a burden for the firm. Since financial markets are imperfectly efficient (Shleifer 2000), the stock market may not value social investments correctly even if it is known that high CSR ratings are good for business. Even if the markets were perfectly efficient (Fama 1970), they may not value social investments correctly because the future is uncertain and it is impossible to predict how a specific social initiative of the company will pay off. As a result, short-term investors may view social investments as risky and uncertain. Long-term investors may be more supportive of social investments than short-term investors due to the long-term payoffs of social investments. Finally, social investments may improve a firm’s reputation, which may allow the firm to differentiate from its rivals and charge premium prices (Shamsie 2003). Such reputation improvements are likely to accrue over time, necessitating a longer time horizon when making social investments.

Although our main proposition in this article is that long-term investors will be more supportive of investments that increase CSR ratings of the firm, we also propose that investors will not support CSR-increasing investments if those investors have other means to profit from the firm. In this regard, we suggest that the institutional context may shape the behavior of the investors. While some countries (i.e. Western economies) may have more transparent markets and more efficient mechanisms of monitoring and control, others may lack such mechanisms or have them in semi-developed, transitional forms. Such a difference in institutional context may affect the relationship between the corporation and its stakeholders. Thus, we would expect to see significant differences among investors in non-Western economies (such as Korea), compared to North America and Europe, regarding their support of CSR.

In sum, the existing evidence suggests that different owners may have divergent, rather than convergent, orientations and preferences regarding corporate strategic decisions. This insight suggests that different owners may also have different orientations and preferences regarding the firm’s strategic decisions about CSR participation. Consequently, the relationship between ownership structure and CSR should vary depending on the shares of the total stock owned by various shareholders. In the subsequent section, we articulate how each type of owner might make distinct decisions on social investments. Since an investigation of the effects of all possible stock owners on CSR is not feasible, we limit the scope of our theorizing to several key owners that have received substantial attention from management scholars, such as institutional and managerial ownership (e.g., Baysinger et al. 1991; Chaganti and Damanpour 1991; Chang 2003; Eisenmann 2002; Hoskisson et al. 2002; Kochhar and David 1996; Zahra 1996). We also include foreign ownership in order to capture the sample-specific characteristic (i.e. large Korean firms) of this study.

Institutional Ownership

Many scholars suggest that institutional owners have significant influence on organizational decisions. For instance, Shleifer and Vishny (1997) argued that institutional owners are influential in organizational decisions by exercising substantial voting power as well as having asymmetric information advantages over other shareholders (Schnatterly et al. 2008). Using their power and information, institutional investors tend to be more actively involved in firms’ decisions than noninstitutional stockholders (Brickley et al. 1988). Moreover, because institutional owners often own significant percentages of the firm’s stock and cannot easily sell their shares, they are likely to be more attentive to the firm’s strategic decisions than other shareholders.

With respect to organization’s decisions on social investments, the good management theory (Graves and Waddock 1994) suggests that the effect of institutional ownership on CSR should be positive. Since the firm’s long-term performance can be enhanced by good management practices, institutional holders are likely to be willing to support CSR-related actions.

An additional rationale explaining why institutional investors might support CSR participation comes from the arguments presented by Siegel and Vitaliano (2007). According to their reasoning, institutional investors such as pension funds, insurance companies, banks, and securities firms offer credence services characterized by significant information asymmetry between the institutional investor and its clients. Investing in socially responsible businesses and maintaining the CSR ratings of the firms is one way for the institutional investor to signal to its potential clients that this institutional investor is reliable and responsible, and thereby to differentiate its services. Following this line of argument, we would expect to see a positive association between institutional ownership and CSR ratings.

Previous studies also support the positive relationship between institutional holdings and CSR. For example, Sethi (2005) argued that public pension funds tend to consider the firm’s long-term effects on the environment, sustainability, and good corporate citizenship when they make an investment decision. Teoh and Shiu (1990) empirically showed that institutional investors look favorably at firms actively engaging in CSR. Graves and Waddock (1994) also noted that institutions invest more heavily in firms with better corporate social performance, finding evidence of a positive relationship between the number of institutions holding the shares of a firm and its CSR rating. Given this description, we predict that institutional ownership will be positively associated with the firm’s engagement in CSR.

H1a

Institutional ownership is positively associated with the CSR rating of a firm.

Even though we assume an overall positive relationship between institutional ownership and CSR, we do not expect that all institutional owners will have the same orientation toward the firm’s social investments, because the incentives of different institutional holders are not always aligned (e.g., Brickley et al. 1988). Thus, institutional investors can fall into different categories since different owners have distinctive motivations and time horizons (e.g., Hoskisson et al. 2002). In this sense, we assume that different types of institutional owners have their own distinct as well as potentially conflicting preferences for the firm’s CSR engagement.

While some institutional investors focus mainly on short-term earnings such as quarterly results (Fortune 1993), others might concentrate on long-term performance. For example, public pension funds have relatively long-term perspectives due to pensioners’ longer time horizons and the sheer size of their holdings. On the contrary, investments by mutual funds, investment banks, security firms, and insurance companies may have relatively short-term orientations due to professional fund managers’ high turnover rates (Bushee 1998) and compensation plans based on short-term earnings (e.g. Johnson and Greening 1999). Since the returns from CSR are expected to be realized mostly in the long run, the long-term oriented institutional investors (i.e., public pension funds) might support CSR, while other short-term oriented institutional investors (i.e., investment banks, securities firms, and insurance companies) may be less supportive of CSR engagement. Therefore, we hypothesize:

H1b

The positive relationship between institutional ownership and CSR rating is stronger for long-term oriented institutional investors than short-term oriented institutional investors.

Managerial Ownership

Agency theory (Eisenhardt 1989; Jensen and Meckling 1976) suggests that top managers have the power to allocate resources among a broad range of stakeholders in a way that assures support from them. However, the theory also suggests that providing stock to managers is an effective way to mitigate agency problems by aligning the interests of the managers with those of the owners. If the managers own significant equity, they are more likely to make decisions maximizing the shareholders’ value (McConnell and Servaes 1990; Denis et al. 1997). If socially responsible actions increase the firm’s value (Orlitzky et al. 2003), as good management theory implies, stock ownership might increase the managers’ incentives to engage in CSR. Empirical findings also support the positive relationship between ownership by managers and CSR engagement. For example, Johnson and Greening (1999) found a positive relationship between top management equity and social performance in terms of environment and product quality.

Another approach to the effect of managerial ownership on the firm’s CSR rating is based on short-term versus long-term orientation. Previously, we have assumed that social performance positively affects firm’s performance. This might be true in the long run but not in the short run because it may be possible for a firm to engage in less CSR and generate higher current profits at the expense of future performance (Narayanan 1985).

In addition, the institutional context in different countries may encourage or discourage socially responsible actions. Developed countries in North America and Europe may impose greater institutional pressures on managers to make socially responsible decisions (Campbell 2007). On the other hand, managers in non-Western countries may not be subjected to similar institutional pressures. In such cases, managers may be more likely to pursue short-term strategies that boost the company’s profits and positively affect their compensation. Lack of transparency in some countries may also create incentives for managers to pursue their personal agendas at the expense of other stakeholders. Arguably, larger stock holdings by the managers in such conditions are likely to endow the managers with greater power to make decisions in their own interests. The financial markets in such economies may not value social investments as highly as in developed economies. If this is true, it would mean that stock-owning managers may not reap the benefits of social investments. Moreover, in some Asian countries (e.g., Japan, China, and Korea) managers of many public firms have strong ties to the founding family (e.g. Claessens et al. 2000), which may wield disproportional influence on the decision-making process. Managers tied to the founding family may adopt policies that benefit the family at the expense of the other stakeholders. Such managers might expropriate value from other shareholders to increase the wealth of the controlling owners such as founding family members (Chang 2003). As a result, the company’s CSR rating may suffer. In this sense, managers in Korea might find more opportunities to profit themselves and the founding family members by reducing social investments than by increasing the long-term value of the company. Overall, we predict that the effect of managerial holdings on CSR ratings will be negative:

H2a

Top management ownership is negatively associated with the CSR rating of the firm.

The resource dependence theory suggests that the selection of outside board members can be seen as a strategic decision for dealing with an organization’s relationships with its business environments (Pfeffer and Salancik 1978). In an attempt to maintain the positive relationships with the environment composed of a wide variety of stakeholders, outside directors may help firms respond appropriately to external constituents through complying with environmental standards and participating in a wide variety of stakeholder-oriented activities (Pfeffer 1973).

However, agency theory (Jensen and Meckling 1976; Fama and Jensen 1983) may suggest that outside directors’ interests are not necessarily aligned with the interests of other shareholders. One mechanism that can be used to align the interests of various parties is the stock ownership. Previous research suggests that outside directors’ ownership of the stock is positively associated with the firm’s financial performance (Hambrick and Jackson 2000). Outside directors that own stock of the firm may be more engaged in monitoring managerial decisions if their own financial wealth is at stake. On the contrary, if outside directors do not own significant shareholdings, they are less effective governance mechanism since their incentive to monitor is low (Gedajlovic and Shapiro 1998). A similar argument may be applied to social investments of the firm. If, as we argued earlier, social investments increase the firm’s long-term value, outside directors owning stock may be inclined to support such investments. In addition, ownership of stock by outside directors may give them additional power and prestige (Finkelstein 1992). All of these factors suggest that outside directors owning a significant percentage of the stock will be inclined to support corporate social investments that increase corporate CSR ratings. Therefore, we hypothesize:

H2b

Outside director ownership is positively associated with the CSR rating of a firm.

Foreign Ownership

It is assumed that higher levels of investment from abroad might indicate a greater influence of foreign practices (e.g., Jeon et al. 2011; Yoshikawa et al. 2010). As an example, the current trends of CSR implementation in many Asian countries have been largely affected by Western-style management practices, which we assume to have higher levels of social engagement. Empirical findings also support this argument. For instance, Chapple and Moon (2005) noted that globalization enhances firms’ CSR engagement in Asian countries. Brancato (1997) also argued that U.S. shareholders have pressured firms to address social responsibility issues for more than 60 years.

A potential caveat of that argument, however, can be found in oversimplifying the attributes of foreign investors and overlooking the variability of their profiles. For example, one may argue that all foreign investors are not always in favor of social investments. Many U.S. and European investment companies have often been involved in antisocial behaviors (e.g., Davis and Kim 2007). Hence, in order to assert the positive influence of foreign ownership on CSR, it is necessary to identify the foreign owners’ profiles that may indicate their investment orientations and preferences.

In the case of Korea, first of all, a substantial proportion of foreign investment has been conducted by countries where CSR is seen as desirable such as Europe and North America. According to the statistics by the Bank of Korea in 2009, foreign investments are largely from Western countries including the EU (32.9%) and the U.S. (20.2%). Since investors from these regional areas might prefer active CSR engagement (Brancato 1997; Chapple and Moon 2005), foreign shareholders from these countries are likely to show similar behavior when they exert their power on Korean firms. For example, Jeon et al. (2011) found that foreign investors with substantial shareholdings (more than 5%) tend to lead Korean firms to pay more dividends (i.e., payout ratio), which is one of the critical evaluative criteria for CSR in Korea.

Second, according to the 2004 annual report of Korea Financial Supervisory Services (KFSS), 99.8% of total foreign shareholdings invested in Korean firms are held by foreign “institutional” investors, including mutual funds (50.9%), banks (20.2%), and pension funds (10.4%), whereas only 0.2% of shareholdings are owned by individual investors. This statistic indicates that foreign ownership in Korea can be best characterized as “institutional” ownership than other kinds.

Furthermore, the reform of Chaebol (i.e. Korean form of business conglomerate meaning “business family”, e.g. Samsung, Hyundai, and LG) has long been a critical issue in Korea, especially since the Asian financial crisis in late 1990s. Old Chaebol firms before the financial crisis were criticized for their “low transparency of corporate decision-making, low accountability of managers, and little or no separation of ownership and control” (Kim 2007, p. 1168). Researchers have argued that foreign investment has a positive impact on the reform the corporate governance of Chaebol (e.g. Haggard et al. 2003), since the reformation was driven largely by the Korean government as a response to criticism from foreign investors. Since the goals of Chaebol reformation (Choi and Aguilera 2009) are aligned with the fundamental ideas of CSR (i.e. emphasizing firms’ social, ethical, and environmental responsibilities), foreign investors are also assumed to have a positive impact on the spread of CSR practices among many Korean firms. Even though Asian firms might not always trail behind their Western counterparts in CSR (Welford 2005), it is assumed that foreign investors can possibly force Korean firms to establish transparent corporate governance and consequently encourage them to engage in CSR to some extent.

Another line of argument relies on the idea of uncertainty reduction that CSR investments may bring. As argued earlier (Siegel and Vitaliano 2007), CSR investments may be a way for companies to differentiate themselves and signal their trustworthiness. Investing in a foreign country is risky and uncertain due to increased information asymmetries (Gehrig 1993). In this case, investing in socially responsible companies is a way to reduce risk for the institutional investor as well as a way to show its clients that the institutional investor itself is highly reputable. Given this line of reasoning, it is rational for foreign investors, especially institutional investors, to invest in socially responsible companies. This line of reasoning does not preclude active participation of foreign investors in decision making. Once the significant investment has been made, the foreign investor will be likely to pressure managers to make socially responsible decisions so as not to lose its investment due to bankruptcy or regulatory/legal sanction. Given the discussions above, we expect that foreign ownership will be associated with higher levels of firms’ CSR ratings.

H3

Foreign ownership is positively associated with the CSR rating of a firm.

Methods

In order to test our hypotheses, we used a sample of large, public Korean firms. As we mentioned, large Korean firms have adapted to the Western-style management practices, especially since the Asian financial crisis in late 1990s. In particular, Korean firms have been paying increasing attention to their stakeholder-oriented practices such as CSR participation. Previous articles found that Korea is one of the few Asian countries that focus on the firms’ social responsibility (e.g., see Chapple and Moon 2005). In this sense, Korea offers an interesting setting for examining the links between ownership and CSR ratings.

Sample

All our sample firms are large Korean firms listed on Korean Stock Exchanges. We initially selected target firms that appear in the list of “2006 top-200 best corporate citizens”, assessed by a leading Korean CSR institution, Korea Economic Justice Institute (KEJI). The CSR ratings are officially labeled as KEJI Index. The trustworthiness of the KEJI Index could be manifested by the 20-year-long history of publication (since its inception in 1991) as well as its extensive usage. The 2006 KEJI Index was based upon the CSR ratings in 2005 (i.e., 2006 is the year of publication).

We also collected firm-level data using the Korea Listed Companies Association’s Directory of Corporate Management (i.e. for more details, see Kim 2005, 2007) and KISVALUE, a Korean electronic database similar to COMPUSTAT in the U.S. Some of the firms in our sample do not have detailed information of the ownership structure such as institutional holdings by each financial institution. Due to the lack of full data availability, our final sample size is 118 firms.

Variables

CSR Ratings (KEJI Index)

KEJI Index relies on multiple distinct data sources to inform the ratings and analysis. It is known that data are collected in a disciplined process from a wide variety of companies, the Korean government, nongovernment organizations, and media sources. Firms are rated with standardized values based on the original interval scales (i.e., A, B, C, D, and E) in seven major sub-domains, including Environment, Community, Corporate Governance, Corporate Integrity, Customer Satisfaction with Product Quality & Safety, Employee Relations, and Long-term Orientation. These domain-specifics appear to be comparable to Kinder, Lydenberg, Domini Research & Analytics (KLD) ratings in the U.S.Footnote 1 Seventeen analysts—six KEJI’s senior analysts and eleven high-profile university faculty with doctoral degrees in economics and business—are known to be involved in creating the KEJI Index. For quality assurance, the KEJI auditing committee, composed of multiple public accountants, performs a quality review of every company profile, updating for content and ratings quality. A full score of KEJI Index is 75. We used 2006 KEJI Index, which represents the CSR ratings in 2005.

Ownership Structure

All ownership data were collected based on both the Korea Listed Companies Association’s Directory of Corporate Management and KISVALUE. We categorized ownership structure into institutional ownership, managerial ownership, and foreign ownership. First, under the umbrella term of institutional ownership, we distinguished four different types of institutional owners in our analysis: (1) public pension funds, (2) insurance firms, (3) securities firms, and (4) investment and commercial banks. There are four major public pension programs in Korea; National Pension, Government Employees Pension, Military Personnel Pension, and Teachers’ Pension (for private school teachers). Institution_Pension Funds is the sum of ownership by all pension funds including these major ones. We also collected the other ownership variables such as Institution_Insurance (ownership by insurance firms), Institution_Securities (ownership by securities firms), and Institution_Banks (ownership by investment/commercial banks). Second, with respect to managerial ownership, we distinguished TMT ownership from Outside Director Ownership. These variables are calculated as the number of shares owned by top managers and outside directors divided by the total number of outstanding shares. Finally, we included Foreign Ownership, which is the percentage of ownership by foreign investors. All ownership information is the average of years 2002, 2003, and 2004. Thus, we have an average 2-year time lag between the measurement of the independent variables (ownership structure) and the measurement of the dependent variable (CSR ratings). This approach to measurement might not entirely establish causality but helps in interpreting the results as an effect of Ownership Structure on CSR, not vice versa.

The descriptive statistics for all shareholders such as means, medians, and standard deviations are reported in Table 1. Overall, pension funds (17.17%) are the most significant institutional shareholder in our sample. Managers (0.11%) and outside directors (less than 0.01%) have relatively small proportions of outstanding shares. Foreign investors (13.71%) also have a significant percentage of ownership in our sample firms.

Table 1 Descriptive statistics of shareholders (%)
Full size table

Control Variables

We included several control variables in order to control for industry and firm characteristics. Firm age, size, financial performance, leverage, and industry dummies were included in the analysis. First, previous research found that Firm Age may be positively (Roberts 1992; Moore 2001) or negatively (Cochran and Wood 1984) associated with the firm’s CSR engagement. Firm age was calculated by the number of years since its foundation. We also measured Firm size by taking the natural logarithm of its total sales, one of the most frequently used measures for firm size (e.g., Hillman et al. 2007; Hambrick and Cannella 2004; Sanders and Boivie 2004). In addition, we controlled for firms’ previous financial performance in order to avoid the alternative explanation by slack-resources theory (Waddock and Graves 1997). Slack-resource theory suggests that because more profitable firms have more organizational slack, they are likely to be more committed to CSR participation. Therefore, we controlled for financial performance by including return on assets (ROA) and leverage by measuring Debt Ratio. Debt ratio was calculated as long-term debt divided by total assets. In order to control for any potential industry effects, we included industry dummies provided by KEJI. Our samples indicate that 45 firms fall into electronics/IT industry, 39 firms for metal/steel/chemistry industry, and 34 firms for food/textile/paper industry. Thus, we created the two industry dummies (metal/steel/chemistry and food/textile/paper) and included them in the analysis (electronics/IT is the reference industry category). All control variables, except industry dummies, are the average over 2002, 2003, and 2004. This again created a time lag so that all our predictive variables are measured as an average of 3 years prior to measuring CSR ratings.

Results

The means, standard deviations, and correlations for the sample firms except industry dummies are presented in Table 2. The average KEJI Index is approximately 47.46 (out of the full score 75) with a standard deviation of 2.46. Correlation results indicate that KEJI Index is positively associated with institutional ownership by pension funds and banks as well as by foreign ownership, whereas it is negatively associated with TMT ownership.

Table 2 Descriptive statistics and correlations
Full size table

Results of OLS regression analysis are reported in Table 3. As the base model, Model 1 has all of the control variables. Firm size (P ≤ 0.001) is positively associated with CSR while high debt ratio (P ≤ 0.05) has a negative relationship with CSR. These findings are consistent with existing theoretical frameworks. For example, the institution-legitimacy perspective (e.g. Stanwick and Stanwick 1998) indicates that institutional pressures often drive larger firms to engage in CSR activities more actively than smaller firms. Moreover, slack-resource theory (e.g. Waddock and Graves 1997) implies that since a high level of debt makes it difficult for a firm to continue to satisfy multiple stakeholders’ expectations, it discourages managers from committing to long-term-focused CSR and forces them to concentrate on increasing the current profits. We also found that firms in food/textile/paper industry were more likely to have higher CSR ratings (P ≤ 0.05), whereas there was no significant difference between firms in metal/steel/chemistry industry and the baseline case of electronics/IT industry.

Table 3 Regression analyses (CSR rating is the dependent variable)
Full size table

Hypotheses for the potential effects of institutional ownership, managerial ownership, and foreign ownership on CSR were tested in Models 2, 3, 4 and 5, respectively. Hypothesis 1a suggested that overall institutional owners will have a positive effect on the firms’ CSR ratings. Model 2 supports this hypothesis: the sum of all institutional holdings (P ≤ 0.01) is positively associated with CSR ratings. In model 3, we tested the effects of each institutional owner on firm’s CSR engagement. Hypothesis 1b suggested that long-term oriented investors may be more likely to support CSR. Our results provide a partial support for Hypothesis 1b: while the long-term investors (i.e. pension funds) have a positive relationship with CSR ratings (P ≤ 0.05), investments by banks also have a positive relationship (P ≤ 0.05) with CSR among the hypothesized short-term investors. Investments by securities firms are marginally significant (P ≤ 0.10) and shareholdings by insurance companies are insignificant. Insurance companies own less than half a percent of stock on average and they invested only 56 firms in our sample. These facts indicate that the effect of ownership by insurance companies is relatively limited. Further interpretation will be elaborated in the discussion section.

Hypothesis 2a stated that managers may be short-term oriented, especially in institutional contexts that do not force managers to invest in CSR and allow managers to pursue their own agendas or founding family’s interests at the expense of other shareholders. Our results in Model 4 support Hypothesis 2a (P ≤ 0.01). Thus, ownership by top managements has a significant negative effect on CSR ratings. However, hypothesis 2b, which assumed a positive relationship between outside director ownership and CSR ratings, is not supported. This finding, however, does not necessarily mean that outside director ownership never has anything to do with CSR ratings. A more convincing reason may be found in the extremely low levels of outside director ownership in Korea (see Table 1). The size of stock holdings by outside directors may not have been sufficient to generate any significant effect.

Finally, hypothesis 3 was tested to examine the relationship between foreign ownership and CSR ratings. As predicted, Model 5 indicates that foreign ownership was found to be positively associated with CSR ratings (P ≤ 0.01). We interpret this result as confirmation of the fact that foreign investors, most of whom come from North America and Europe, may place a relatively greater emphasis on socially responsible business practices. The reason would be their institutional norms at home, which usually encourage social responsibility of business. Foreign investors are also likely to be long-term oriented, given the fact that most of them are institutional investors. Finally, foreign investors will be likely to pressure Korean firms to adopt socially responsible practices because of their desire to signal to their clients that these investors are reliable and responsible firms.

We do not test the full model with all independent variables at once for the following reasons: (a) our variables might not be mutually exclusive and (b) there is a possible lack of independence among the owners’ decisions. First, each ownership variable can be overlapped (e.g. double counting). For instance, an investment by a European pension fund might be regarded as both an institutional holding and foreign ownership at the same time. Second, a potential problem with our dataset is the possible lack of independence among various investors’ decisions. For example, it is possible that foreign investors may experience more severe information asymmetry problems. In that case they might imitate Korean institutional investors’ decisions. Another possibility is that Korean institutional investors may follow the lead of large foreign investors. Either way, including the unmodified variables for institutional ownership and foreign ownership in the same equation would violate the assumption of OLS regression about independence of residuals.

In order to correct these issues, we used 2-stage least squares (2SLS) regression, putting all of the variables in the same equation. We performed the first stage analysis by regressing foreign investment on the other independent and control variables and then using the residuals as the new foreign ownership variable in the second stage OLS regression, thereby controlling for any effects that institutional investors may have had on the decisions made by foreign investors. The same technique was used for the opposite effect of foreign ownership on institutional ownership. Two sets of 2SLS regressions generated similar results with our original OLS regression: a positive effects of ownership by institutions and foreign investors on CSR as well as negative effects of top manager holdings on CSRFootnote 2.

In sum, all key owners significantly explain more variance in CSR ratings as one can see in the R-square change over the control model; overall institutional holders (7.5%), managerial ownership (7.2%), and foreign ownership (5.6%). This indicates that ownership structure has significant effects on the firm’s CSR engagement. Specifically, ownership by institutions including pension funds, investment/commercial banks, and securities firm (albeit only marginal) are positively associated with the firm’s CSR rating. In addition, foreign ownership, mainly from the countries where CSR is well established such as the U.S. and Europe, is also positively related to the CSR rating. Ownership by top managers is negatively associated with CSR engagement and outside director ownership is not significant.

Discussion

Based on the argument that shareholders have different motivations and time horizons for corporate decisions (e.g., Hoskisson et al. 2002), we tested if different owners have distinct influences on the firm’s CSR engagement. We found that, as predicted, investors with long-term orientation such as large institutional shareholders and foreign investors support the firm’s CSR initiatives. However, top managers were found not to support active engagement in CSR. We will discuss the findings based on existing theoretical frameworks and the Korean context.

Effects of Institutional Ownership on CSR

Our results show that ownership by large institutions such as pension funds and banks is positively associated with CSR ratings. Institutional shareholders holding significant firm equity cannot easily sell their stocks without severely lowering the stock price (e.g. Pound 1992), so their investments are likely to be based upon the long-term competitiveness of the firm. We also assumed that certain types of institutions (i.e. insurance companies, banks, and securities firms) may pay more attention to short-term performance, which induces less investment in CSR. However, our results suggest that ownership by banks and securities firms (albeit marginal) is also positively associated with CSR ratings.

There are a few reasons for this finding. First, even though financial investors such as banks and securities firms may be under more pressure for immediate economic gain than pension funds, these institutional owners may be unable to easily divest their shares without significantly lowering the stock price. Accordingly, they cannot be purely short-term oriented if they have significant levels of ownership. Second, banks and securities firms tend to consider not only potential returns but also financial risks when they make investment decisions. Investment in a firm that is socially irresponsible is not efficient (Graves and Waddock 1994) because socially irresponsible firms are exposed to the higher risk of regulatory action, legal punishment, or consumer activism. In this sense, banks and securities firms might achieve similar return with less risk by investing in firms with higher social performance.

Finally, the Korean government bailed many financial institutions out in a process of recovering from the Asian financial crisis. For example, the government took over Woori Finance Group, one of the largest financial groups in Korea, in 1998 after injecting 12.8 trillion won (i.e. 10.8 billion U.S. dollars) in public funds to bail out Woori Bank and its peers that later consolidated into a holding company. Thus, financial institutions are not independent from the influence of the Korean government, which has emphasized the importance of institutions’ social responsibility as a financial resource provider, rather than a short-term economic gain seeker. Since financial institutions are under the government influence, it is hard for them to over-emphasize economic gain at the expense of social responsibility.

Effects of Managerial Ownership on CSR

In addition, we tested the effects of managerial ownership by top managers and outside directors on CSR ratings. Agency theory implies that managerial ownership can align potentially divergent interests of the shareholders and managers. If CSR increases the firm’s value, as the good management theory (Graves and Waddock 1994) suggests, managerial ownership leads to more active engagement in CSR. In contrast with this prediction, our findings indicate that ownership by Korean top managers is negatively associated with CSR ratings. The negative relationship between managerial ownership and CSR could be understood in the Korean context. Korean large firms including Chaebol have been characterized by “little or no separation of ownership and control” (Kim 2007). It indicates that Chaebol family members and relatives, who also own significant amount of shares, frequently serve as managers of the firm or exert significant influence on managers. They have been criticized for misbehavior; for instance, a study of 250 Korean firms belonging to 38 family-run conglomerates found that a quarter of them revealed irregular transaction records aiming at enriching the Chaebol family (Albrecht et al. 2010; Moon 2006). Chang (2003), in a study of 419 Chaebol affiliates firms during the period 1986–1996, also found that larger shareholders who are also managers are able to expropriate value from other shareholders. Thus, managers from the Chaebol family with significant shares are likely to overuse their power given by the formal title and ownership. These behaviors are against the firm’s ethical and social responsibility. In this sense, ownership by managers in Korea might be associated with decisions only enriching the family or relatives of Chaebol even at the expense of other public stakeholders, in turn, reducing the firm’s CSR ratings.

Furthermore, it is surprising that managers have a relatively small proportion of share (mean = 0.11%), but exert significant influence on CSR ratings. Even though their stockholdings are relatively small, they might have more organizational decision power by complex ownership characteristics such as “pyramid structures” and “cross-holdings” (for detail see Chang 2003; Claessens et al. 2000). In this case, decision rights or voting rights frequently exceed the rights formally entitled by ownership size. Chaebol (i.e. business groups), usually consists of a number of diversified firms, which is similar to multidivisional organization, under which individual companies function as operating divisions. Since individual firms are often under influence of group-level staff (Chang & Hong, 2000) or controlling family, managers might exert more power beyond their entitled ownership, especially when they are related to or controlled by the founding family of Chaebol. The complex ownership structure and influence of Chaebol might explain the significant negative effects of managerial ownership even though they have such a small proportion of shareholdings.

Moreover, we did not find evidence between outside director ownership and CSR ratings. First, the majority of Chaebol affiliates’ boardrooms are generally filled with insiders and friends of Chaebol families (Chang 2003, p. 241). Thus, outside directors might not encourage managers, who are not favorable to firm’s social responsibility, to engage in socially responsible actions. Second, more importantly, the size of stock holdings by outside directors may not have been sufficient to generate any significant effect. In our Korean firm sample, most of our sample firms do not report any outside director ownership and the reported maximum outside director ownership is only 0.02%. These facts imply that the effect of director ownership in our sample is highly limited. Finally, since powerful large shareholders such as Chaebol have resisted the governance reform such as empowering outside directors (e.g. Cho and Kim 2007), the role of outside directors is relatively restricted in Korea, especially when they do not possess sufficient ownership. Previous literature (Gedajlovic and Shapiro 1998; Pedersen and Thomsen 1997) found that the role of outside directors is limited for those companies in which there is lack of separation of ownership and control.

Effects of Foreign Ownership on CSR

We found that the higher levels of investment into Korea from abroad are associated with enhanced CSR ratings. Chapple and Moon (2005) argued that “the higher the level of investment into a country from abroad, the higher the likely influence of foreign practices on domestic companies (p. 420)”. Since the current trends in CSR are influenced by Western practices, the level of Western investments is assumed to be related to CSR engagement. In particular, investments in Korea from abroad are largely from Western countries such as the E.U. (32.9%) and the U.S. (20.2%), so these investors are likely to facilitate CSR participation. As previously argued, since the Asian crisis of late 1990s, Korean firms have been making more efforts toward achieving responsible and transparent governance structures. Foreign investors have played an important role in this process by shareholder activism and board participation (Choi et al. 2007). In addition, as we argued earlier, foreign investors may use CSR ratings as a guide to making investments due to significant information asymmetries between foreign investors and Korean companies. Having made the investment based on the CSR ratings, foreign investors may start to pressure Korean managers to further improve their companies’ social performance.

Conclusion

We found that ownership structure has significant effects on the firm’s CSR engagement. This study makes at least two significant contributions. First of all, there have been relatively few studies (e.g. Barnea and Rubin 2010; Johnson and Greening 1999) on the relationship between ownership structure and CSR. Our study reconfirms the argument of previous studies that ownership structure affects strategic decisions of the firm (Baysinger et al. 1991; Hart and Moore 1990; Kochhar and David 1996) by showing that investors have different attitudes toward CSR engagement. In particular, we categorize owners into three separate groups: institutional (banks, pension funds, insurance companies, and securities firms), managerial (top management team and outside directors), and foreign ownerships. No previous CSR research investigated the effect of such detailed ownership structure on CSR. We found that top managers in Korea tend to be less interested in improving their firms’ CSR ratings than institutional investors and foreign owners. The results enhance our understanding of the relationship between ownership structure and CSR.

Second, this study examined the relationship between ownership structure and CSR in a non-Western context. Most of the previous studies on CSR have largely been conducted in an institutional environment where the CSR is relatively well established such as Western Europe and the U.S. (Belal 2001). However, recent studies pointed out that corporations elsewhere in the world other than Europe and the U.S., have recently adopted CSR practices (Chapple and Moon 2005; Matten and Moon 2008). By examining CSR in non-Western context, we found that the ownership–CSR relationship has a distinct implication in Non-Western contexts as compared to Western contexts. Using a sample of large Korean firms, our study reveals that foreign investors played a significant role in the adoption of Western-style management practices like CSR engagement. In addition, although financial institutions are highly motivated by short-term economic gain, ownership of financial institutions influenced by the government affects the firm’s CSR participation differently. Our results imply that financial institutions are motivated to encourage CSR if they are under the influence of the government, as is the case in Korea. This study shows that the relationship between ownership structure and CSR can vary by the institutional context. Jones (1999) also emphasized the importance of socio-cultural environment and national institutional arrangement on CSR. In this sense, we can draw the conclusion that the dynamics of an institutional environment affect the firm’s strategic choices (Peng 2003).

In addition to these theoretical contributions, there are some policy and practical implications that can be drawn from our findings. The policy implications are as follows. First, this article supports the legitimacy of the governance reform which has been implemented by the Korean government since late 1990s. As we have shown, excessive power of the dominant coalition such as the Chaebol family is detrimental to CSR engagement. In order to promote CSR engagement, the government might need to monitor the ownership structure of the dominant coalition, which is associated with the decision rights. Second, our findings provide insightful implications for the government’s foreign ownership restriction policy. Numerous countries have adopted ownership restriction policies toward foreign investors, especially in emerging economies such as China (e.g. Bailey and Jagtiani 1994; Chen et al. 2001). However, the previous literature found that foreign ownership is positively associated with financial performance (e.g. Greenaway 2009) and sustainability (e.g. Kronborg and Thomsen 2009). Additionally, Choi et al. (2007) noted that foreign investors, who entered Korean stock market after the abolition of the foreign equity restriction policy in 1998, played a significant role as external governance mechanisms. Our finding sheds additional light on the positive effects of foreign ownership. Allowing foreign owners to own more of the domestic companies might facilitate socially responsible management and social investments.

Our findings also provide practical implications. First, organizations need to be careful to administer any stock-based compensation to managers and directors. Our findings suggest that there may be a detrimental effect of inside ownership in terms of CSR engagement, especially when there is a dominant coalition (e.g. the Chaebol family). Second, our findings show that different owners may have divergent and even competing perspectives on CSR. For instance, top management owners have a negative viewpoint toward CSR, whereas institutional and foreign owners have a positive outlook. Such competing strategic preferences may incur inefficient procedures on social investment decisions. Hence, firms need to develop more sophisticated intervention mechanisms (e.g., an introduction of a CSR-focused division and executive officers; namely, CRO) to deal with conflicting voices that may produce procedural complications.

In spite of these contributions, this study is not without limitation. First, our study was done in one institutional context (Korea), therefore, generalization of our results should be done with caution. It is possible that replication of this study in other countries would produce different results. For example, the negative effect of managerial ownership on CSR ratings may also be partially attributed to the Korean context, especially the family orientation of Chaebol. Future research needs to extend the study’s context by investigating the ownership–CSR association in other Non-Western countries. Second, our data were collected cross sectionally, which means that we cannot claim causation (even though there is a two-year average lag between the measurements of the independent and dependent variables). In addition, due to the lack of longitudinal data, we cannot tell how stable the relationships between ownership structure and CSR ratings are across time. Future research needs to conduct longitudinal examinations to validate the findings of this study. Third, this study does not explore the dynamic interactions among owners in terms of CSR. Drawn from existing literature focusing on the interactions among different shareholders regarding various corporate outcomes (e.g., Chaganti and Damanpour 1991), future research needs to investigate the dynamics among different types of owners regarding CSR.

In conclusion, to date CSR has received substantial attentions from both academic scholars and practitioners. Among others, factors and conditions that determine CSR have been of a primary interest. This study advances the existing knowledge by proposing and providing evidence that different owners have distinctive impacts on CSR participation. It is our hope that this insight can have a meaningful impact on the stream of research connecting corporate governance to CSR.