Another Form of Greenwashing: The Effects of Chaebol Firms’ Corporate Governance Performance on the Donations

Abstract

Environmental, social, and governance (ESG) metrics are widely used to measure the firms’ social performance. In this regard, donation expenses are one of the outcomes for the firms’ effort to build, grow, and maintain the social value. However, firms may expense a trivial or minimum amount of donations, considering the corporate size, in order to disguise themselves as a “good company”. In this paper, exploiting 2010–2019 Korean Stock Exchange (KSE) market listed companies’ financials and ESG scores, we examine whether Chaebol firms with good governance “actually” spend more donation expenses. We predict and find that good governance does not actually lead to greater donation expenses among Chaebol firms, despite the positive relations between governance and donation expenses in general. Overall, our findings highlight that good Chaebol companies determined by ESG metrics may not be real charitable companies. Our findings provide counterevidence against the notion that firms with a higher ESG score are more likely to be charitable.

1. Introduction

Charitable donations are one of the most important social works of a firm, because they build positive corporate images, help promote the products and services of a firm, and lead to an increase in the firm’s value. Ultimately, in the long term, charitable donations contribute to the sustainable growth of society. However, considering the socially and financially irresponsible behaviors of Chaebols such as aggressive tax avoidance, accounting fraud, and scandals, we expect that not all the charitable donations of Chaebol firms are genuine.

The purpose of our study is to investigate whether charitable donations of Chaebol firms are actual and genuine. More to the point, we examine the relation between Chaebols, corporate governance, and charitable donations. We examine the financial data of a sample of Korean Stock Market listed companies that also have data on KCGS dataset of environmental, social, and governance (ESG) ratings of publicly traded companies. To identify firms with better corporate governance, we construct a CG variable for the period 2010 through 2019.

We find that Chaebols are positively related to charitable donations, suggesting that Chaebol firms are more likely to spend on charitable donations than non-Chaebol firms. Given that Chaebol firms often enjoy competitive advantages relative to non-Chaebol competitors, available resources for donations from Chaebol firms are bigger than non-Chaebol firms and it leads to the bigger giving.

We also find that better governance is positively related to charitable donations, suggesting that better governance increases charitable donations. The result is consistent with the argument that charitable donations increase a firm’s value when the governance mechanism of the firm is strong enough to limit the private interest of the management and major shareholders; whereas, charitable donations decrease a firm’s value when the governance mechanism of the firm is too weak to monitor the management and major shareholders (Kim and Kang, 2014 [1]).

Lastly, we find that governance is negatively related to charitable donations given by Chaebol firms. It suggests that charitable donations of Chaebol firms would focus on advertisement and investment attraction, rather than focus on actual and socially meaningful charitable activities.

Our results make three important contributions to research on ESG and donations. First, our results help in understanding the negative relation between Chaebol firms’ governance and donations. Second, our results provide policy implications that regulators consider extra tax credits for genuine and respectable amounts of charitable donations. Third, our results also remind benevolent consumers and investors who believe that a firm’s charitable donations are important should interpret information on charitable donations carefully and critically.

The remainder of this paper is structured as follows: Section 2 provides an overview of backgrounds and develops hypotheses. Section 3 introduces research design. Section 4 reports empirical results and Section 5 concludes.

2. Backgrounds and Hypotheses Developments

There are four streams of prior literature: motivations and effects of charitable donations, corporate governance and charitable donations, reactions of investors to the charitable donations, and greenwashing.

First, to the extent of motivations of charitable donations, Navarro (1988) [2] reports that corporate contributions to charitable actions are likely to be driven by profit and utility maximization motives. He utilizes advertisement expenditure, labor cost percentage, sales margin, debt ratio, dividend change, managerial compensation, and effective tax rate and finds charitable donations can be seen as a kind of advertisement. In a similar vein, Boatsman and Gupta (1996) [3] estimate firm-specific marginal tax rates for a five-year period spanning the Tax Reform Act of 1986 (TRA86) and find negative relations between the marginal tax rate and charitable donations. They contend that utility maximization of managers, rather than profit maximization of firms, is an important motive for corporate charitable donations. Brammer and Millington (2005) [4] report that firms with higher levels of philanthropic expenditures are likely to have better reputations. They contend that charitable donations can create more positive impressions, therefore firms may make use of them for stakeholder management. Lev et al. (2010) [5] find a positively significant relation between charitable donations and future revenues. More particularly, they find the positive association between charitable donations and future revenue is pronounced when consumer perception is important. Kim et al. (2017) [6] find that corporate size, market share, and brand value lead to greater charitable donations among the listed companies in Korea. Xia et al. (2019) [7] report that Chinese firms are likely to increase charitable donations after regulatory penalties. They find the positive relation between regulatory penalties and charitable donations are more pronounced when the firms are non-state owned, are highly concentrated in ownership, and are experiencing severer penalties.

Second, prior studies document the relation between corporate governance structure and charitable donations. Barnea and Rubin (2010) [8] argue that corporate social responsibility expenditures may be consistent with a value maximization goal when it comes to the response to stakeholders’ preference changes, while managers may benefit from overinvestment in CSR for their good reputations. They find insider ownership and financial leverage are negatively related with a CSR score, implying that motivations for overinvestment in CSR grow when insiders carry little or no cost of CSR.

Kim et al. (2017) [6] state that charitable donations are likely to be driven by the private benefit of insiders rather than a strategic accumulation of intangible assets such as reputations, i.e., firms with high-powered managers are inclined to spend more on charity. They find that foreign ownership and an ESG score are positively related with charitable donations, implying that foreign investors are likely to believe corporate philanthropic practices lead to increased long-term profit and survivability.

Hong (2019) [9] documents that charitable donations from Chaebol firms have the propensity to decrease shareholder value, while donation activities enhance shareholder value in general, implying that the corporate philanthropic practices of Chaebol firms have plenty of room for improvement.

Third, some researchers focus on the investors’ reaction to charitable donations. Brown et al. (2006) [10] find that an involvement in giving increases shareholder value, especially when competitors in the same industry are likely to be involved in similar giving programs.

Maung et al. (2019) [11] find a more positive relation between charitable donations of family business with religion-declared CEOs and firm value. Their findings are consistent with the notion that donation activities may enhance moral capital and shareholder value, only if those are not perceived as self-serving actions (Godfrey, 2005 [12]).

Chen et al. (2020) [13] document that investors have the propensity to exaggerate corporate ESG information, i.e., investors express positive responses to good news about firms with higher ESG scores yet negative responses to bad news about firms with lower ESG scores In addition, they argue that due to the market participants’ overreaction to ESG information a sophisticated investor can enjoy abnormal returns that exploit ESG momentum strategy.

Fourth, prior literature documents the drivers, conditions, and market reactions to greenwashing behavior. Delmas and Burbano (2011) [14] contend that uncertainty about regulatory punishment for greenwashing behavior may contribute to greenwashing. Wu et al. (2020) [15] develop the analytical models that connect greenwashing and information transparency. They argue that low information transparency motivates profit-driven firms to engage in greenwashing, while high information transparency eliminates greenwashing behavior. They also find high information transparency may encourage socially responsible firms to make extra explicit corporate socially responsible (CSR) investments, including donations. Du (2015) [16] documents that greenwashing is negatively related to stock returns, while corporate environmental performance is positively related to stock returns in China. The author argues that market participants tend to value environmentally genuine firms. More directly, Gatti et al. (2021) [17] report that greenwashing, which refers to one of the deceptive communications, has a negative impact on investment intentions.

Unfortunately, to date, research directly addressing the relation between charitable donation and greenwashing are rare. However, several donation activities are directly or indirectly connected with greenwashing behavior, especially when a company’s donation strategy focuses on environmental campaigns. For example, companies perform cause-related marketing, which refers to a marketing tool that promotes product purchasing that leads to target-oriented donations. Considering that corporate donations toward environmental protection have been remarkably increasing [18,19] and market participants are more likely to value the generous giver [9], environmentally poor performers and givers would engage in greenwashing to meet the social and market demands.

Some studies argue that corporate philanthropic practices are driven by non-financial motivations such as earning reputations (Brammer and Millington, 2005 [4]). Given that Chaebol firms refer to large business groups and play a significant role in Korea’s economy, it is necessary for Chaebol firms to manage their reputations to make good impressions on the public and to avoid political costs. If this is the case, Chaebol firms are eager to increase their charitable donations. However, other studies contend that private benefits of insiders are the main motivations for philanthropic expenditures (Kim et al., 2017 [6]). In that case, agency problems and corporate governance structures play a crucial role in determining the amount of charitable donations. These predictions lead to our first and second hypothesis:

Hypothesis (H1).

Chaebol firms are more likely to donate to charities.

Hypothesis (H2).

Firms with better governance are more likely to donate to charities.

According to Parguel et al. (2011) [20], sustainability ratings might deter greenwashing, because higher sustainability ratings are likely to lead to positive corporate image and vice versa. Additionally, politically vulnerable firms have the propensity to increase charitable donations and investors tend to overreact to ESG information. If this is the case, Chaebols might put genuine efforts on charitable donations that are higher than average.

On the contrary, firms might spend a minimum number of philanthropic expenditures and exaggerate their charitable activities, even if it is not genuine. We can borrow ideas from cause-related marketing, which is defined as the cooperative marketing campaign of a company or non-profit organization. For example, Krombacher, a German brewery company, started a rainforest protection campaign where the company directly donated 5 cents per crate of beer to the World Wildlife Fund between 2002 and 2008. However, people accused the Krombacher campaign of not donating enough in relation to the size and the publicity of the campaign (Lütge, 2018 [21]).

In this regard, given that several Chaebol firms or their founding family members have carried out socially irresponsible practices such as accounting frauds [22], aggressive tax avoidance [23], and scandals [24,25], it is possible for Chaebol firms to be involved in greenwashing by making use of charitable donations. More specifically, Chaebol firms might enjoy good reputations due to participation in charitable activities with minimum cost. These discussions lead to our third hypothesis:

Hypothesis (H3).

Chaebol firms with good governance are less likely to donate to charities.

3. Research Design

3.1. Measuring Variables

3.1.1. Identifying Chaebol Firms

The Chaebol refers to a business conglomerate system of South Korea, which accounts for a large part of the Korean economy. To the extent of corporate governance, Chaebols are usually controlled and managed by a family member of the group founder. The Korea Fair Trade Commission (KFTC) defines Chaebol as “a group of companies of which the contents of business are mostly controlled by the same entity”. More specifically, if a firm fulfills either the shareholding ratio or the control requirement, it is classified as Chaebol affiliates.

Table 1. Criteria for Chaebol Classification.

Requirement	Criteria
Shareholding ratio	- A company in which the same person, acting alone or together with any of the following persons or entities (hereinafter referred to as “person related to the same person”), holds at least 30 percent of the total number of shares issued by that company (excluding non-voting shares) as the largest investor of that company.
Control	- A company, the representative director of which is appointed or dismissed by, or at least 50 percent of executives of which has been or can be appointed by the same person under contract or agreement with other major shareholders. - A company subject to controlling influence by the same person, directly or through a person related to the same person, with respect to major decision-makings or business execution such as reorganization of the relevant company and investments in new business. - A company that exchanges its personnel exchanges with the relevant company and a company controlled by the same person (including the same person, where the same person is a company). - A company that makes transactions of funds, assets, goods, services, etc., with the same person or a person related to the same person company, or that gives and takes debt guarantees to or from the same person or a person related to the same person company in excess of the ordinary scale, or any other company recognized as an economic entity identified with the same person according to social norms since the company conducts its business in a way that it can be considered as an affiliate of the business group of the same person.

Source: Korea Fair Trade Commission (www.egroup.go.kr (accessed on 22 October 2021)) and Enforcement Decree of the Monopoly Regulation and Fair-Trade Act (law.go.kr (accessed on 22 October 2021)).

reports the criteria for Chaebol classification in detail.

3.1.2. Measuring Corporate Governance Performance

Corporate governance performance consists of shareholders’ rights protection, characteristics of board, disclosure, audit function, and dividend payment. The Korea Corporate Governance Service (KCGS) announces governance ratings as part of annual ESG reporting. The value of CG ranges 1 to 6: CG denotes 6, when the governance rating is A+. On the contrary, CG denotes 1, when the governance rating is D. Thus, a greater CG value indicates a greater level of governance rating.

3.1.3. Measuring Charitable Donations

Charitable donations refer to cash and non-cash items made to nonprofit organizations such as foundations and NGOs to help them accomplish their goals. The purpose of charitable donations from a firm is to enhance its brand image and reputation and to deliver social impact. Mostly, those charitable donations made by a firm are viewed as benevolent and altruistic actions. However, in cause-related marketing studies, charitable action is interpreted as self-interest rather than altruism when a firm benefits first before any commitment to donate is accrued (Dean, 2003 [26]). The variable DON, proxy for charitable donations of a firm, is calculated as the amount of donation expenses divided by the beginning total assets and multiplied by 1000 (Hong, 2019 [9]).

3.2. Empirical Model

To empirically test H1, H2, and H3 we use the following regression model (1). The dependent variable, DON, is the level of charitable donations. Among independent variables, we expect the Chaebol dummy (CB) and the corporate governance rating (CG) to determine the level of charitable donations. We also include an interaction term between the two variables (CB × CG) to determine whether the effect of governance on charitable donations varies with the type of the company (Chaebol vs. non-Chaebol).

Testing the H1, the regression coefficient of CB captures the relation between Chaebol and charitable donations. A positive coefficient would suggest that Chaebol firms have a tendency to spend more on charities than non-Chaebol firms, whereas a negative coefficient would suggest that Chaebol firms are less likely to spend on charities. With respect to the H2, the coefficient of CG captures the relation between governance and charitable donations. A positive coefficient would suggest that firms with better governance have the propensity to spend more on charities than firms with worse governance. However, a negative coefficient would suggest that firms with good governance are less likely to spend on charities. Lastly, testing the H3, our variable of interest is the CB and CG interaction, which captures the differential (Chaebol vs. non-Chaebol) charitable donation expense per governance rating. If Chaebol firms with better governance have lower (higher) charitable donations, then the CB × CG coefficient should be negative (positive).

We include control variables that prior literature finds to be important determinants of charitable donations. Firm size (SIZE) and leverage (LEV) are included, based on Hardwick and Adams (2002) [27], who document the positive (negative) relation between firm size (leverage) and charitable donations and cash flow from operations (CFO) and sales growth (GW) control for the capacity to generate cash flow and growth opportunity, respectively. The return on assets (ROA) is also based on Hardwick and Adams (2002) [27], who find that greater profitability is positively associated with the amount of charitable donations. The loss indicator variable (LOSS) is included as a control for firms reporting negative earnings. Advertisement expense (ADV) is also included, because the amount of branding expense is associated with donation strategy. Firm age (AGE) and foreign ownership (FOR) are included as controls for underlying reputations and ownership structure of a company, respectively. In addition, the regression model has year and industry fixed effects, where industry is based on the KSIC middle industry classifications.

DON = b₀ + b₁CB + b₂CG + b₃CB × CG + b₄SIZE + b₅LEV + b₆CFO + b₇GW + b₈ROA + b₉LOSS + b₁₀ADV + b₁₁AGE + b₁₂FOR + FixedEffects + e

3.3. Data and Sample Selection

Our sample consists of Korean Stock Exchange listed firms, for which data are available on DART and TS2000 for the years 2010 through 2019. Data on the ESG ratings, which we use to construct our governance measure, come from the KCGS (Korea Corporate Governance Service) dataset of environmental, social, and governance ratings of listed companies. Our final sample consists of 6218 firm–year observations.

4. Results

4.1. Descriptive Statistics

Table 2. Descriptive Statistics (n = 6218).

Variable 1	Mean	Std.	Min.	Median	Max.
CB	0.197	0.398	0.000	0.000	1.000
CG	2.959	0.902	1.000	3.000	6.000
DON	0.860	1.995	0.000	0.176	18.177
SIZE	19.952	1.485	16.277	19.714	23.930
LEV	0.407	0.206	0.037	0.410	0.933
CFO	0.046	0.074	−0.207	0.042	0.311
GW	0.052	0.250	−0.609	0.028	1.273
ROA	0.024	0.075	−0.277	0.026	0.240
LOSS	0.224	0.417	0.000	0.000	1.000
ADV	0.617	1.440	0.000	0.058	11.589
AGE	3.501	0.717	0.693	3.738	4.407
FOR	10.070	12.738	0.000	4.590	55.990

¹ Variable definitions: CB–equal to 1 if the firm is included in the Chaebol category, and 0 otherwise; CG–corporate governance performance, equal to 6 if the firm–year earned A+ rating, equal to 1, if the firm–year earned D rating, on the contrary; DON–donation expenses × 1000 scaled by beginning total assets; Tobin’s q–market value of assets divided by book value of assets; MTB–market-to-book value ratio; LEV–total liabilities to total assets; SIZE–the natural logarithm of total assets; CFO–cash flows from operations scaled by beginning total assets; GW–growth rates in sales; ROA–return on assets; LOSS–equal to 1 if the firm reported negative net income, and 0 otherwise; ADV–advertisement expenses scaled by beginning total assets; AGE–the natural logarithm of firm age; FOR–foreign ownership ratio, which is number of shares held by foreign investor divided by total number of shares outstanding.

presents the descriptive statistics for the variables used in our hypotheses tests. The mean of CB is 0.197, indicating that approximately 19.7 percent of our samples are classified as Chaebol firms. The maximum (minimum) CG is 6.0 (1.0), which refers to A+ (D) governance rating and the governance score of the median sample firm is 3. Mean DON is 0.860, representing that the average firm spends approximately 0.086 percent of its total asset on charitable donations per year. The most charitable firm in our sample spends 18.177 percent of its total asset on donations.

4.2. Univariate Analyses

Table 3. Mean Difference of Donation Expenses (n = 6218).

Panel A: Chaebol vs. Non-Chaebol Samples
Chaebol (n = 1228)	Non-Chaebol (n = 4990)	Diff(a-b)	t-stat
1.129	0.794	0.335	5.29 ***
Panel B: Better vs. Worse Governance Samples
Better (n = 4772)	Worse (n = 1446)	Diff(a-b)	t-stat
0.752	0.145	0.607	7.80 ***

*** indicates statistically significant at level 1%.

reports test results for the mean difference of charitable donation expenses. Panel A presents the mean difference between Chaebol and non-Chaebol groups. The 1228 Chaebol firm–years (Mean = 1.129) compared to the 4990 non-Chaebol firm–years (Mean = 0.752) demonstrates a significantly higher level of donation expenses to total assets (Diff. = 0.335; t = 5.29 ***). Panel B presents the mean difference between better and worse governance groups. The result indicates that firms with better governance (Mean = 0.752) spend more on charitable donations, while firms with worse governance (Mean = 0.145) spend less on charitable donations. The mean difference between the two groups is statistically significant (Diff. = 0.607; t = 7.80 ***).

4.3. Correlations

Table 4. Correlation Matrix (n = 6218).

Variables	CB	CG	DON	SIZE	LEV	CFO	GW	ROA	LOSS	ADV	AGE
CG	0.376 ***
DON	0.067 ***	0.104 ***
SIZE	0.573 ***	0.487 ***	0.067 ***
LEV	0.121 ***	−0.019	−0.086 ***	0.164 ***
CFO	0.086 ***	0.161 ***	0.174 ***	0.190 ***	−0.121 ***
GW	−0.006	0.016	0.034 ***	0.015	0.027 **	0.099 ***
ROA	0.050 ***	0.165 ***	0.194 ***	0.180 ***	−0.295 ***	0.506 ***	0.199 ***
LOSS	−0.047 ***	−0.138 ***	−0.117 ***	−0.134 ***	0.292 ***	−0.344 ***	−0.171 ***	−0.683 ***
ADV	−0.026 **	0.089 ***	0.204 ***	0.058 ***	−0.081 ***	0.133 ***	0.050 ***	0.128 ***	−0.079 ***
AGE	−0.029 **	−0.097 ***	−0.086 ***	−0.031 **	−0.013	−0.086 ***	−0.110 ***	−0.073 ***	0.033 ***	−0.084 ***
FOR	0.287 ***	0.351 ***	0.137 ***	0.527 ***	−0.135 ***	0.284 ***	0.015	0.262 ***	−0.172 ***	0.198 ***	−0.059 ***

Pearson correlation coefficients are reported. ** and *** indicate statistically significant at level 5% and 1%, respectively. See Table 1 for variables definitions.

documents the Pearson correlation coefficients between each of the measures exploited in our study. With respect to the variables of interest, we find CB and CG are positively and significantly correlated with DON. Comparing the control variables, we find SIZE, CFO, GW, ROA, ADV, and FOR are positively and significantly correlated with DON, whereas LEV, LOSS, and AGE are negatively and significantly correlated with DON.

4.4. Regression Analyses

Table 5. Regression Results for H1, H2, and H3 (n = 6218).

Variables	(1)		(2)		(3)
	Coefficient	t-Value	Coefficient	t-Value	Coefficient	t-Value
Intercept	2.983	6.38 ***	2.514	5.99 ***	2.898	6.14 ***
CB	0.230	3.07 ***			0.753	3.35 ***
CG			0.099	3.16 ***	0.136	3.71 ***
CB × CG					−0.163	−2.61 ***
SIZE	−0.038	−2.47 **	−0.029	−1.34	−0.056	−2.34 **
LEV	−0.340	4.58 ***	−0.316	−2.30 **	−0.322	−2.34 **
CFO	1.756	−1.25	1.731	4.52 ***	1.699	4.43 ***
GW	−0.123	8.10 ***	−0.119	−1.21	−0.112	−1.14
ROA	3.949	2.02 **	3.870	7.95 ***	3.895	7.99 ***
LOSS	0.160	8.45 ***	0.165	2.09 **	0.172	2.18 **
ADV	0.159	−3.67 ***	0.154	8.21 ***	0.156	8.30 ***
AGE	−0.125	3.59 ***	−0.116	−3.41 ***	−0.116	−3.39 ***
FOR	0.009	−2.47 **	0.008	3.33 ***	0.008	3.50 ***
YD	Included		Included		Included
IND	Included		Included		Included
Adj. R2	0.137		0.137		0.139
F-value	29.12 ***		29.14 ***		27.99 ***
Sample size	6218		6218		6218

Table 5 reports testing results for H1, H2, and H3. All dependent variables are DON (donation expenses). ** and *** indicate statistically significant at level 5% and 1%, respectively. See Table 1 for variables definitions. Full regression model for column (3): DON_t = β₀ + β₁CB_t + β₂CG_t + β₃CB × CG_t + β₄SIZE_t + β₅LEV_t + β₆CFO_t + β₇GW_t + β₈ROA_t + β₉LOSS_t + β₁₀ADV_t + β₁₁AGE_t + β₁₂FOR_t + ΣYD + ΣIND + ε_t.

reports testing results for H1, H2, and H3. Column (1) of Table 5 presents the result of estimating Equation (1) when the primary variable of interest is CB. The regression coefficients of the primary variable of interest CB is positive and significant (Coeff. = 0.230; t = 3.07 ***). It suggests that consistent with our prediction, Chaebol firms have a tendency to spend more on charitable donations than non-Chaebol firms. Given that the revenue and the size of Chaebol firms are bigger than those of non-Chaebol firms, on average Chaebols are more likely to be capable of being a ‘big giver’ than non-Chaebol firms.

Column (2) presents the testing result for H2, when the primary variable of interest is CG. We find a significant and positive relation between CG and DON (Coeff. = 0.99; t = 3.16 ***). This result suggests that firms with better corporate governance are likely to spend more on charitable donations. In other words, on average governance has the propensity to increase the ratio of donation expenses on total assets, implying that a better governance structure leads to high levels of corporate philanthropy, whereas worse governance is likely to restrain charitable donations. Also, it is similar to the argument of Bartkus et al. (2002) [28] that some influential stockholders may perceive donations as excessive spending and some governance mechanisms cut off corporate philanthropy.

Lastly, column (3) reports the negative regression coefficient on CB × CG (Coeff. = −0.163; t = −2.61 ***). This result suggests that Chaebol firms with a better corporate governance structure, surprisingly tend to decrease charitable donations relative to non-Chaebol firms. It implies that Chaebol firms would spend charitable donations at a certain point, exaggerate their charitable actions much more than actual, and thus mislead consumers and investors. Results on the control variables are economically modest.

5. Additional Analyses

Endogeneity Issues

We report the negative and significant relation between CB × CG and DON (charitable donations) in Table 5. However, some missing variables might influence both donations and Chaebol. To resolve endogeneity problems: (1) we estimate a two-stage least squares (2SLS) regression model, and (2) we exploit the propensity score matching (PSM) method to construct matched samples from our original dataset and again estimate the relation between CB × CG and DON.

To the extent of the 2SLS estimation, we carefully select share turnover (Turnover), which proxies for information asymmetry and transparency as an instrumental variable, because corporate governance literature argues that information transparency is correlated with Chaebol. More specifically, Chaebol firms have a propensity to exhibit lower information transparency than non-Chaebol firms [29].

Table 6. Robustness Test Results: 2SLS Regression with Instrumental Variable (n = 6218).

Variables	(1) First Stage Dependent Variable = CB		(2) Second Stage Dependent Variable = DON
	Coefficient	T-Value	Coefficient	z-Value
Intercept	−0.084	−3.23 ***	2.912	4.44 ***
CB			10.929	2.73 ***
Turnover	0.006	5.12 ***
CG	−0.071	−37.82 ***	0.858	2.99 ***
CB × CG	0.262	222.11 ***	−2.810	−2.67 ***
SIZE	0.012	7.69 ***	−0.226	−3.33 ***
LEV	0.022	2.77 ***	−0.429	−2.72 ***
CFO	0.062	2.84 ***	1.672	3.26 ***
GW	−0.009	−1.61	−0.002	−0.02
ROA	0.003	0.11	3.230	5.60 ***
LOSS	−0.006	−1.36	0.262	2.71 ***
ADV	−0.001	−0.71	0.207	10.35 ***
AGE	−0.001	−0.40	−0.132	−3.34 ***
FOR	−0.001	−4.35 ***	0.012	3.48 ***
YD	Included		Included
IND	Included		Included
Adj. R2	0.929		n/a
F-value	3239.23 ***		n/a
Wald Chi2			516.25 ***
Sample size	6218		6218

*** indicates statistically significant at level 1%.

reports 2SLS estimation results for the main hypothesis, H3. We find a significant and negative relation between CB × CG and DON (Coeff. = −2.810; t = −2.67 ***) from the second stage estimation, suggesting that the main results of our study are found to be robust on the potential endogeneity issues.

Also, the PSM method is widely used in accounting and finance literature to alleviate selection bias from observables and endogeneity problems [30,31]. To examine whether our main results are robust on potential endogeneity and reverse causality issues, we construct a one-to-one (Chaebol vs. non-Chaebol) matched sample using the PSM method and re-estimate our regression model.

Table 7. Robustness Test Results: One-to-One Matched Sample using PSM approach (n = 2456).

Variables	Dependent Variable = DON
	Coefficient	t-Value
Intercept	4.645	4.86 ***
CB	1.084	3.02 ***
CG	0.400	5.23 ***
CB × CG	−0.363	−3.83 ***
SIZE	−0.226	−4.94 ***
LEV	−0.817	−2.94 ***
CFO	0.823	1.03
GW	−0.160	−0.77
ROA	11.305	10.91 ***
LOSS	0.599	3.85 ***
ADV	0.292	7.05 ***
AGE	0.058	0.88
FOR	0.003	0.72
YD	Included
IND	Included
Adj. R2	0.306
F-value	30.31 ***
Sample size	2456

*** indicates statistically significant at level 1%.

documents the estimation results for H3 within the artificially constructed one-to-one matched sample with the PSM method. Consistent with the main results, the regression coefficient on CB × CG is found to be negative and significant (Coeff. = −0.363; t = −3.83 ***). It suggests that the original results remain robust on the selection bias and endogeneity issues. Overall, the additional test results lead to the same conclusion as the primary test results.

6. Conclusions

We investigate the relation between Chaebol, corporate governance, and charitable donations. Our findings are as follows:

First, we find positive and significant relations between Chaebol and charitable donations. It suggests that on average Chaebol firms are more likely to play a ‘big giver’ role than non-Chaebol firms, because the size and available resources of Chaebol firms are bigger than non-Chaebol firms.

Second, we also find positive and significant relations between corporate governance and charitable donations, which suggests that better corporate governance leads to more charitable donations. This evidence is consistent with the argument that charitable donations increase firm value when the governance mechanism of the firm is strong enough to limit the private interest of the management and major shareholders (Kim and Kang, 2014 [1]).

Third, we find negative and significant relations between Chaebol and corporate governance interaction terms (CB × CG) and charitable donations; suggesting that Chaebol firms may spend charitable donations provided that the donation activities are well advertised. In other words, charitable donations of Chaebol firms would focus on advertisement and investment attraction rather than emphasize actual and socially innovative charitable actions. Considering the socially and financially irresponsible behaviors of several Chaebol firms, our evidence implies that some may exaggerate their charitable actions much more than the actual, and thus mislead consumers and investors.

Lastly, we also perform a 2SLS estimation and a PSM approach to investigate whether our main results are robust on the potential endogeneity problems; the results from the additional analyses suggest that our main conclusions are robust on the endogeneity issues.

Our results have implications for those stakeholders who view corporate philanthropy as an important part of a firm’s activities. Particularly, benevolent consumers and investors need to pay attention to interpreting Chaebol firms’ governance and donations. More to the point, regarding the policy implications, regulators may consider additional tax credits for an actual and respectable amount of charitable donations. On the contrary, as Delmas and Burbano (2011) [14] stated, limited and imperfect information about regulatory punishment may contribute to greenwashing. Therefore, the government should prepare for guidelines and decrease uncertainty about punishment for engaging in deceptive behaviors that exploit charitable donations. Managers and practitioners of companies that are similar to the business and governance structure of Chaebol (i.e., family-owned conglomerates) must objectively evaluate whether their donation activities are authentic and must keep improving social performance. We hope the results of the paper will be a step toward a richer and more inclusive understanding of the reality of charitable donation activities.

In environmental, social, and governance studies, greenwashing refers to disguising consumers and investors through exaggerating environmental efforts and pretending to be an environmentally sound company. Our evidence suggests that charitable donations of Chaebol firms, even firms with better governance mechanisms, would not be genuine charity. It would be another form or extension of greenwashing that impedes sustainable growth of the economy.