Corporate environmental responsibility (CER) weakness, media coverage, and corporate philanthropy: Evidence from China

Abstract

This study investigates the hidden connection between corporate philanthropy and corporate environmental responsibility (CER) weakness. Using a sample of Chinese listed firms in polluting industries and hand-collected data on corporate environmental performance and corporate philanthropy, we show that CER weakness is significantly positively associated with corporate philanthropy, suggesting that corporate philanthropy may be used by environmentally unfriendly firms to mitigate the negative influence of CER weakness and offset pressures from stakeholders. This finding also implies that Chinese enterprises in polluting industries are inclined to engage in greenwashing via the conduit of corporate philanthropy. In addition, media coverage reinforces the positive association between CER weakness and corporate philanthropy. Above results are still valid after controlling for the potential endogeneity between CER weakness and corporate philanthropy.

Corporate social responsibility (CSR) is a multi-dimensional concept and different CSR dimensions may be inherently inconsistent (Chen, Patten, & Roberts, 2008). Extant studies have documented strategic, altruistic, political, and managerial self-interested motivations of corporate philanthropy, as well as the determinants and the economic consequence of corporate philanthropy (Du, 2015b; Porter & Kramer, 2002; Saiia, Carroll, & Buchholtz, 2003; Zhang, Zhu, Yue, & Zhu, 2010), but provide insufficient evidence on the internal relationships among different CSR dimensions. This study fills the above gap by investigating whether corporate environmental responsibility (CER) weakness influences corporate philanthropy and examining the moderating effects of media coverage.

Environmental conservation is a worldwide focus. In China, the second largest economy and the largest developing country in the world, environmental conservation is particularly urgent. In 2006, China became the world’s largest emitter of greenhouse gases, the giant evil of global warming (Du, Jian, Zeng, & Du, 2014). In fact, environmental destruction has become commonly seen phenomena in China and a number of firms greedily grab for profit at the expense of environmental deterioration. However, for a long time, officials turn a blind eye to environmental wrongdoers for the sake of GDP growth and political promotion. As a result, environment pollution has not been effectively controlled. For the past few years, public awareness of environmental protection is increasingly strengthened, so polluting firms in China are confronted with unprecedented pressures. As a response, “greenwashing” becomes a common tactic that is used by polluting enterprises in China to offset pressures from stakeholders. Greenwashing always means a firm’s poor environmental performance and positive communication about its environmental performance simultaneously (Delmas & Cuerel Burbano, 2011), suggesting that a firm manipulates its environmental performance to mislead consumers, investors, and other stakeholders (Du, 2015a). Specifically, many firms carry out corporate philanthropy to divert the public’s attention away from the negative image of their environmental wrongdoings (Du, 2015b). As a result, the public can observe the freak phenomenon that corporate philanthropy is associated with environmental saboteurs and wrongdoings. In this regard, the public feel very puzzled about the real motivation of corporate philanthropy in Chinese enterprises and further question whether corporate philanthropy is used for the purpose of covering environmental wrongdoings. Motivated by above issues, this study examines the hidden connection between corporate philanthropy and CER weakness.

Using hand-collected data on corporate philanthropy, corporate environmental performance, and media coverage, we examine whether CER weakness influences corporate philanthropy and further investigate the moderating effect of media coverage. For empirical tests, we construct a sample of Chinese listed firms from polluting industries over the period of 2009–2011. In brief, our findings show that CER weakness is significantly positively associated with corporate philanthropy, suggesting that corporate philanthropy may be used by firms from polluting industries to mitigate the negative influence of environmentally unfriendly behavior and divert the attentions of the public away from negative environmental events. Moreover, media coverage reinforces the positive association between CER weakness and corporate philanthropy.

Our study contributes to the existing literature in several ways. First, this study is one of very few studies to investigate the hidden connection between corporate philanthropy and CER weakness. Second, this study validates CER weakness dressing (i.e., greenwashing) as an additional motivation of corporate philanthropy. Third, this study recognizes the positive influence of media coverage on corporate philanthropy. Fourth, this study explores the reinforced effect of media coverage on the positive association between CER weakness and corporate philanthropy. Fifth, using a sample of firms from polluting industries in China, our study provides supplementary evidence to previous literature based on the context of developed markets. Finally, our study is one of very few studies to calculate CER weakness based on Global Reporting Initiative (GRI) guidance, a multi-dimensional evaluation index system.

Literature review, theory, and hypotheses development

Literature review on the framework of factors influencing corporate philanthropy

Extant studies have paid close attention to the determinants and consequences of corporate philanthropy. Specifically, scholars have recognized that internal governance mechanisms (e.g., blockholder ownership, the ratio of independent directors, board size, the percentage of shares held by directors/top managers, board gender diversity, and institutional investors, etc.), firm-specific financial characteristics (e.g., firm size, financial leverage, cash holding, accounting performance, and growth opportunity, etc.), and the nature of the ultimate owner impact corporate philanthropy (Brammer & Millington, 2005; Chen et al., 2008; Crampton & Patten, 2008; File & Prince, 1998; Gao, Faff, & Navissi, 2012; Maas & Liket, 2011; Mitschow, 2000; Zhang et al., 2010).

In addition, for recent years, researchers also argue and validate that external monitoring or governance mechanisms (e.g., analyst coverage, product market competition, government intervention, industry competition, and political forces, etc.) influence corporate philanthropy (Chapple & Moon, 2005; Fernández-Kranz & Santaló, 2010; Johnson, 1966; Li & Zhang, 2010). Harjoto and Jo (2011) found the positive association between CEOs’ turnover and corporate philanthropy. Fernández-Kranz and Santaló (2010) found that firms in more competitive industries carry out higher level of corporate philanthropy. Li and Zhang (2010) found that political interference should be responsible for a firm’s philanthropy. Chapple and Moon (2005) documented that globalization enhances CSR in Asian countries. Johnson (1966) provided the analysis framework to argue that both government intervention and industry regulation influence corporate philanthropy.

Furthermore, a branch of very thin but growing literature (Chen et al., 2008; Delmas & Cuerel Burbano, 2011; Du, 2015a; Koehn & Ueng, 2010; Laufer, 2003; Prior, Surroca, & Tribó, 2008; Williams & Barrett, 2000) addresses the impacts of various corporate wrongdoings on corporate philanthropy, in which corporate philanthropy is used by these wrongdoers to offset the negative images on corporate operations, suggesting that firms may use corporate philanthropy as moral window-dressing. For example, Prior et al. (2008) documented that firms experiencing earnings management are inclined to resort to CSR practices. Chen et al. (2008) argued that corporate philanthropy may cover up misconducts in low quality products, employee discrimination, and food security, among others. Williams and Barrett (2000) found that firms that violate regulations carry out corporate philanthropy to save their suffered reputations. Koehn and Ueng (2010) found that corporate philanthropy may be used by accounting wrongdoers to transfer investors’ attentions from the negative influence of the financial restatement. Moreover, Delmas and Cuerel Burbano (2011) and Laufer (2003) suggested that greenwashing is likely to be associated with environmental window-dressing. Du (2015a) found that the market negatively reacts to a firm’s greenwashing. As the response, a greenwashing firm may engage in corporate philanthropy to cover environmentally irresponsible behavior, and thus greenwashing becomes more invisible and inconspicuous.¹

Institutional background and hypotheses development

In fact, accompanying with the rapid economic growth in China (one of the biggest “world’s factories”), the environmental problems are deteriorating (Baskin, 2006). Why does environmental conservation in China have difficulty in taking a step? First, local governments have political and economic motivations to deregulate on environmental conservation for seeking GDP one-sidedly (Du, 2015b; Lin,2010), resulting in underdeveloped environmental responsibility. Moreover, local government officials are inclined to be short-sighted. That is, officials always focus on nominal economic development but ignore pollution problems accompanying with industrialization (Küskü, 2007; Mansden, 2000). As a response, firms in polluting industries always greedily pursue profit maximization at the expense of reckless environmental destruction. Second, there is no statutory policy to compel firms to fulfill their environmental responsibilities. Third, for a long time, there has not been an independent and efficient judicial system and thus the existing laws, regulations, and rules about environmental protection are performed less effectively. As a result, environmental protection is far from being optimistic and laws and regulations are just on paper (Du, 2015b). Finally, business ethics are still under construction, and thereof many firms refuse to pay for environmental protection because of lacking the intrinsic consciousness of environmental responsibility (Du et al., 2014).

Nevertheless, ironically, many heavily polluting enterprises such as Zijin Mining Co. (601899.SH) often appear on the charity lists in China. In addition, tobacco enterprises and papermaking enterprises are always recommended as candidates for various China Charity Awards. The phenomenon, that is, a firm carries out philanthropy on a public occasion but actually is an environmental wrongdoer, motivates us to question whether the real motivation of corporate philanthropy is altruistic (Campbell, Gulas, & Gruca, 1999) or is just environmental misconduct dressing or greenwashing (Du, 2015b; Lin, 2010).

In fact, Zyglidopoulos, Georgiadis, Carroll, and Siegel (2012) argued that CSR includes CSR-strength and CSR-weakness. CSR-strength refers to “the additional benefits beyond those required by law and narrow economic interest that a firm provides to its stakeholders” and CSR-weakness means “the negative effects that a firm’s operation has on its stakeholders that remain after a firm’s CSR activities” (Zyglidopoulos et al., 2012: 1623). In this regard, researchers can classify corporate philanthropy as one of CSR-strengths and recognize CER as one of CSR-weaknesses. Zyglidopoulos et al.’s (2012) classification can help us discuss the internal relationships among various dimensions of CSR, especially, corporate philanthropy and CER.

Using the Chinese context, Du (2015b) found that a firm carries out corporate philanthropy but wrecks the environment simultaneously. As such, corporate philanthropy does not always mean a virtue or an ethical behavior. When Du’s (2015b) findings are applied to Zyglidopoulos et al.’s (2012) classification between CSR-strength and CSR-weakness, we can draw a tentative conclusion that a firm may cover up CSR-weaknesses through CSR-strengths. However, using survey data, Du (2015b) based his study on environmentally illegal activities (some extreme cases) and thus the findings may not fit in well with the association between CER weakness and corporate philanthropy. Caulkin (2002) sharply criticized the typical phenomena in which a firm carries out corporate philanthropy on the one hand but destroys environment on the other hand. Caulkin (2002) and Du (2015b) argued that environmental wrongdoers carry out corporate philanthropy just for the purpose of covering up their environmentally unfriendly behavior or/and to transfer the attentions of the public away from negative images of environmental wrongdoing. Moreover, Chen et al. (2008) and Koehn and Ueng (2010) also found that some firms carry out corporate philanthropy to alleviate the negative influence of deficiencies in other dimensions of CSR.

Theoretically, although a variety of other CSR dimensions may be used by firms to cover their CER weaknesses or offset the negative influence of environmental wrongdoings, Zyglidopoulos et al. (2012) suggested that corporate philanthropy is a channel or conduit with relatively lower costs compared with other CSR dimensions. Therefore, focusing on the internal relationship between CER weakness and corporate philanthropy, we predict the positive association between CER weakness and corporate philanthropy and formulate the following Hypothesis 1 in an alterative form:

Hypothesis 1

Ceteris paribus, CER weakness is positively associated with corporate philanthropy.

Zyglidopoulos et al. (2012) recognized that media coverage strengthens CSR. Moreover, extant studies find that stakeholders positively value a philanthropic firm, which results in better performance, lower costs of debts, higher firm value, and even better corporate image (Goss & Roberts, 2009; Margolis & Walsh, 2003; Orlitzky, Schmidt, & Rynes, 2003; Zhang et al., 2010). As a result, it is rational for a firm with higher media coverage to carry out corporate philanthropy to offset stakeholder pressures.

Furthermore, we discuss the moderating effect of media coverage on the association between CER weakness and corporate philanthropy. Media coverage always “combines public and non-public information with an analysis that highlights potential problems” (Miller, 2006: 1006) to serve as a monitor or “watchdog” (Djankov, McLiesh, Nenova, & Shleifer, 2002), and thus can early identify corporate environmental misconduct or CER weakness. In fact, “the press can serve as an information intermediary and thus potentially shape firms’ information environments by packaging and disseminating information, as well as by creating new information through journalism activities” (Bushee, Core, Guay, & Hamm, 2010: 1). Especially, the nongovernment-owned presses per se are market participants which can ensure themselves to undertake original analysis. As a result, extant studies (Bragdon & Marlin, 1972; Dasgupta, Laplante, & Mamingi, 2001; Du, 2015a; Hamilton, 1995; Huberman & Regev, 2001; Klassen & McLaughlin, 1996) document the influence of media coverage on the market reactions to CER weakness. For example, Du (2015a) found that firms in the list of greenwashing have significantly negative cumulative abnormal returns (CAR) compared with environmentally friendly firms.

In contemporary China, accompanying with aggravating environment, the press increasingly pay their attentions to CER weakness. To alleviate the negative market reaction to CER weakness, a firm always “attempts to generate goodwill in the eyes of the public through corporate philanthropy” and seeks to “buy off” officials and regulators with a view to make them more lenient through corporate philanthropy (Du, 2015b: 341). Also, firms with CER weaknesses have a higher likelihood of receiving media attention and thus have a higher probability of exposure to risk (Godfrey, 2005). Therefore, media coverage compels firms with CER weaknesses to carry out corporate philanthropy to mitigate the negative influence from environmental wrongdoings or misconducts.

Above discussions, taken together, motivate us to predict that the positive association between CER weakness and corporate philanthropy is more pronounced for firms with strong media coverage than for firms with weak media coverage. Therefore, we formulate the following Hypothesis 2:

Hypothesis 2

Ceteris paribus, media coverage reinforces the positive association between CER weakness and corporate philanthropy.

Methodology

Sample

The initial sample consists of all Chinese A-share firms in polluting industries (wood and furniture, papermaking and printing, metal and non-metal, machinery, equipment, and instrument manufacturing, other manufacturing, construction, and transportation and warehousing) for the period of 2009–2011. We begin with 2207 observations, and then select the sample in the light of the following criteria: (1) We delete firm-year observations with negative net assets or shareholders equity (363 observations). (2) We exclude firm-year observations that have listed for less than 1 year (58 observations). (3) We discard firm-year observations whose data on firm-specific control variables are unavailable (188 observations). Finally, we obtain a sample of 1598 firm-year observations. Then, the top and bottom 1 % of each variable’s distribution are winsorized to alleviate the influence of extreme observations.²

The data sources for variables in this study are as below: (1) We collect the data of GIVING based on CSMAR (China Stock Market and Accounting Research) database, a frequently used database in extant China studies. (2) We calculate the data of CER_WEAK based on hand-collected data of the raw score of corporate environmental disclosure. (3) We hand collect and calculate data on MEDIA from the website of http://news.baidu.com. (4) We hand collect data on POL_CON and MON. (5) We obtain data on BIG10 and GOV_INT from the website of www.cicpa.org.cn and Fan, Wang, and Zhu (2010), respectively. (6) Other data are all calculated and collected based on CSMAR.

Corporate philanthropy (dependent variable)

Corporate philanthropy (cash and goods donations) is reported in a sub-item of non-operating expenses in Chinese listed firms’ financial statements (Zhang et al., 2010). In general, there are four kinds of measures for corporate philanthropy: (1) the amount of corporate philanthropy deflated by sales revenue (Chen et al., 2008); (2) the amount of corporate philanthropy deflated by total assets at the beginning of the year (Brown, Helland, & Smith, 2006); (3) the amount of corporate philanthropy deflated by pre-tax income/profit in the year (McGuire, Sundgren, & Schneeweis, 1988; Roberts, 1992; Wang & Coffey, 1992; Ullmann, 1985); and (4) the natural logarithm of the amount of corporate philanthropy (Zhang et al., 2010). In this study, we adopt the first method to measure GIVING (the dependent variable) as the amount of corporate philanthropic giving deflated by sales revenue.

Corporate environmental responsibility weakness (main independent variable)

CER weakness (CER_WEAK) is the inverse proxy for corporate environmental responsibility. To control the potential endogeneity between CER weakness and corporate philanthropy, we lag one period for CER weakness and employ CER_WEAK _t−1 as the main independent variable.

In this study, based on the GRI sustainability reporting guidelines (GRI, 2006), we employ content analysis to conduct the scoring procedure by extracting environmental information from every firm-year’s annual report, CSR report, and other disclosure. Specifically, corporate environmental disclosure includes seven components: governance structure and management systems, credibility, environmental performance indicators, environmental spending, vision and strategy claims, environmental profile, and environmental initiatives. These components can further be divided into 44 subcomponents according to Clarkson, Li, Richardson, and Vasvari (2008). Moreover, based on scores of 44 subcomponents and seven components (see Appendix for details), we calculate the raw score of corporate environment disclosure. Finally, we convert the raw score of corporate environment disclosure into CER weakness (i.e., CER_WEAK), measured as “(Max_ENV-ENV _{j, t} )/ (Max_ ENV−Min_ ENV)”. ENV _{j, t} is the corporate environmental performance for firm j in year t. Max_ ENV and Min_ ENV are the maximum and the minimum values of ENV _{j, t} by industry and year, respectively. In our study, if the coefficient on CER_WEAK is significantly positive, Hypothesis 1 is supported by empirical evidence.

Media coverage (moderating variable)

The popular approach to collect data on media coverage (MEDIA) in the existing literature is to calculate the number of a firm’s name appearing in headlines of news reports in paper media (Fang & Peress, 2009). However, data based on paper media may face the risk of underestimating the intensity of media coverage because people are increasingly inclined to obtain information from the internet, in addition to paper media. Moreover, paper media may fail to exactly measure the number of information dissemination, which embodies the true value of news in the media (press) to a great extent. In this study, with the help of “Baidu News Search Engine” (http://news.baidu.com), we calculate the total number of annual news reports about each firm by identifying whether the headlines of Baidu news cover a firm’s name, stock name, or stock code. ³ Furthermore, we measure media coverage (with the label of MEDIA) as the natural logarithm of (1 + the number of media coverage). This measurement better captures information dissemination including almost all news websites, websites of newspapers, magazines, broadcasting, and TV, and thus it is less likely to underestimate the intensity of media coverage. However, we can not judge the tendentiousness one by one due to the great capacity of Baidu news regarding our approach of measuring media coverage. Consistent with Zyglidopoulos et al. (2012), the coefficient on MEDIA is expected to be significantly positive. Also, if the coefficient on CER_WEAK×MEDIA is positive and significant, Hypothesis 2 is validated.

Control variables⁴

First, to control the influence of internal corporate governance on corporate philanthropy, we include FIRST, BD_SHR, INST_SHR, DUAL, INDR, and BOARD into our models (Johnson & Greening, 1999). FIRST is the percentage of common shares held by the controlling shareholder. BD_SHR denotes the percentage of shares owned by directors. INST_SHR is the percentage of shares owned by institutional investors. DUAL is an indicator variable, equaling 1 if the CEO and the chairman of the board are the same person and 0 otherwise. INDR denotes the ratio of independent directors, measured as the number of independent directors to the total number of the board of directors. BOARD is the natural logarithm of the number of directors in the boardroom (Wang & Coffey, 1992).

Second, following extant literature, we include four variables of firm-specific financial characteristics into our models. SIZE is the natural logarithm of total assets (Brammer & Millington, 2006; Meznar & Nigh, 1995; Useem, 1988). LEV is the debt-to-asset ratio, measured as total liabilities deflated by total assets (Brown et al., 2006). OCF is net operating cash flow deflated by total assets. TOBIN’Q is a firm’s market value divided by total assets at the end of the year (Clarkson et al., 2008).

Third, as the response to extant studies (Chapple & Moon, 2005; Fernández-Kranz & Santaló, 2010; Johnson, 1966; Li & Zhang, 2010), we include a set of variables about external governance mechanisms such as ANALYST, BIG10, CEO_CH, CMP, POL_CON, INTER, MON, GOV_INT, IND_REG, and CROSS in our regressions. ANALYST is the natural logarithm of (1 + the number of analyst coverage). BIG10 is a dummy variable, equaling 1 if a firm’s auditor is a Big 10 accounting firm (including affiliated firms) according to the official rank of the Chinese Institute of Certified Public Accountants and 0 otherwise. CEO_CH is an indicator variable, equaling 1 if a firm changes its CEO and 0 otherwise. CMP denotes product market competition, measured as the Herfindahl index of a firm’s sales revenues in various industries. POL_CON is a dummy variable, equaling 1 if a firm’s CEO is politically connected and 0 otherwise (Du, 2015b). INTER is the ratio of a firm’s overseas sales including sales from exporting and foreign subsidiaries to its total worldwide sales (Geringer, Tallman, & Olsen, 2000; Hitt, Hoskisson, & Kim, 1997). MON denotes monitoring intensity, measured as the distance between a firm and the nearest regulatory center (Beijing, Shanghai, and Shenzhen) divided by 1000 (El Ghoul, Guedhami, Ni, Pittman, & Saadi, 2013). GOV_INT is the government intervention index from Fan et al. (2010). IND_REG is a dummy variable, equaling 1 if a firm is located in a highly-regulated industry. CROSS is an indicator variable, equaling 1 if a firm’s stock is listed in two or more stock markets and 0 otherwise.

Fourth, we incorporate DAC, SAL_DIS, PENALTY, and MISSTATE into our study to control for the hidden motivation of corporate philanthropy (Chen et al., 2008; Du, 2015b; Koehn & Ueng, 2010; Prior et al., 2008). Therefore, DAC is the absolute value of discretional accruals following Kothari, Leone, and Wasley (2005). SAL_DIS denotes the ratio of the average pay of top managers to the average wage of employees excluding top managers. PENALTY is the number of penalties that a firm was punished by the China Securities Regulatory Commission. MISSTATE is the amount of income restatement scaled by total assets at the beginning of a calendar year (Files, Swanson, & Tse, 2009).

Fifth, we also introduce STATE into our regression models (Zhang et al., 2010). STATE is a dummy variable, equaling 1 when the controlling shareholder of a firm is a government agency or government controlled state-owned enterprise and 0 otherwise.

Finally, we also include industry and year dummy variables (fixed effects) into our regressions.

Results

Descriptive statistics and Pearson correlation analysis

Section A of Table 1 presents the descriptive statistics of variables used in this study. As shown in Section A, the mean value of GIVING (multiplied by 1000), the dependent variable, is .247. In addition, the distribution of GIVING does not obey standard normal distribution, suggesting that this study should adopt Tobit regression procedure. The mean value of CER_WEAK is .893, suggesting that CER weakness (the inverse proxy for corporate environmental performance) is very serious for Chinese listed firms from polluting industries.⁵ The mean value of MEDIA is 6.244, meaning that the average number of media coverage is about 514 [e^6.244−1]. Moreover, the descriptive statistics of control variables (see Section A of Table 1 for details) are reasonably distributed.

Table 1 Descriptive statistics and Pearson correlation matrix

Section B of Table 1 reports Pearson correlation analysis, and all correlation coefficients equal to or greater than the absolute value of .041, .049, and .064 are statistically significant at 10, 5, and 1 % level, respectively. CER_WEAK is significantly positively associated with GIVING at the 1 % level, providing preliminary support to Hypothesis 1. Moreover, there is a marginally significantly positive relation between MEDIA and GIVING, which can borrow support from Zyglidopoulos et al. (2012). Above results, taken together, suggest the need to address the moderating effect of MEDIA on the relation between CER weakness and corporate philanthropy.

As for Pearson correlation between GIVING and control variables, we find that GIVING is significantly positively related with BD_SHR, DUAL, SAL_DIS, and MON, but is significantly negatively associated with FIRST, ANALYST, BIG10, SIZE, GOV_INT, IND_REG, CROSS, and STATE. These results, taken together, suggest a need to control for these variables when we examine the influence of CER weakness on corporate philanthropy. Moreover, as expected, the coefficients of pair-wise correlation among control variables are generally low, suggesting no serious multicollinearity problem exists when these variables are included in the regression simultaneously.

Multivariate regression results

Table 2 reports the step-by-step Tobit regression results of GIVING on CER_WEAK, MEDIA, and other determinants. All reported t-statistics are based on robust standard errors adjusted for clustering at firm level and year level (Petersen, 2009). Furthermore, from Column (1) to Column (4) from Model (1) to Model (4), as shown in the bottom line in Table 2, four step-by-step Tobit regressions display gradually increasing explanatory powers with significantly higher pseudo R ² values.

Table 2 Regression results of corporate philanthropy on CER weakness, media coverage, and other determinants

Column (1) and Model (1) in Table 2 report the regression results of corporate philanthropy on all control variables. Results show that corporate philanthropy (GIVING) is significantly positively associated with BD_SHR, INST_SHR, DUAL, BOARD, SIZE, DAC, SAL_DIS, PENALTY, and MON, but is significantly negatively related with ANALYST, BIG10, INDR, MISSTATE, IND_REG, CROSS, and STATE, echoing findings in extant literature or consistent with theoretical expectations (Chapple & Moon, 2005; Chen et al., 2008; Koehn & Ueng, 2010; Williams & Barrett, 2000).

Hypothesis 1 predicts that CER weakness (CER_WEAK) is positively associated with corporate philanthropy (GIVING). Column (2) and Model (2) in Table 2 report regression results of GIVING on CER_WEAK and other determinants. As shown in Column (2), the coefficient on CER_WEAK is positive and significant at the 1 % level (.3017 with t = 2.92), providing strong support to Hypothesis 1. Moreover, this result also reveals the following aspects: First, this result can borrow some support from extant studies (Caulkin, 2002; Chen et al., 2008; Du, 2015b; Koehn & Ueng, 2010). These previous studies find that corporate philanthropy is used by various wrongdoers (e.g., accounting, production, and employee relations wrongdoer, etc.) to divert stakeholders’ attentions away from the corresponding negative events. Second, the hidden motivation of corporate philanthropy in environmentally unfriendly firms from polluting industries may be wrongdoing-dressing, rather than a virtue. Finally, the coefficient estimate indicates that one standard deviation increase in CER weakness can increase corporate philanthropy by 4.68 % (.3017 × .155), equivalent to about 18.95 % of the average corporate philanthropy (4.68 %/.247). Clearly, this coefficient estimate is economically significant.

Column (3) and Model (3) in Table 2 report Tobit regression results of GIVING on CER_WEAK, MEDIA, and other determinants. The coefficient on CER_WEAK is positive and significant at the 1 % level (.3037 with t = 2.88), suggesting that Hypothesis 1 is still valid after considering the influence of media coverage on corporate philanthropy. In addition, MEDIA has a significantly positive coefficient (.0434 with t = 4.12), revealing that media coverage can upgrade the level of corporate philanthropy.

Hypothesis 2 predicts that media coverage reinforces the positive association between CER weakness and corporate philanthropy. Column (4) and Model (4) in Table 2 report Tobit regression results of GIVING on CER_WEAK, MEDIA, the interaction between CER weakness and media coverage (CER_WEAK×MEDIA), and other determinants. As shown in Column (4), the coefficient on CER_WEAK is significantly positive (.2265 with t = 2.24), lending additional and strong support to Hypothesis 1. Moreover, the coefficient on MEDIA is positive and significant at the 1 % level (.0359 with t = 3.33), consistent with findings in Column (3). More importantly, as shown in Column (4) and Model (4), the coefficient on CER_WEAK×MEDIA is positive and significant at the 1 % level (.2149 with t = 2.60), providing strong support to Hypothesis 2 and suggesting that media coverage reinforces the positive association between CER weakness and corporate philanthropy.

Next, we use Fig. 1 to show the interaction between CER weakness (CER_WEAK) and media coverage (MEDIA) on corporate philanthropy (GIVING). In Fig. 1, the black, red, and blue lines denote the influence of CER weakness on corporate philanthropy for the full sample, the low-MEDIA subsample, and the high-MEDIA subsample, respectively. First, as Fig. 1 displays, the positive association between CER weakness and corporate philanthropy is more pronounced for the high-MEDIA subsample (blue line) than for low-MEDIA subsample (red line). Second, Fig. 1 shows the positive and negative associations between CER weakness and corporate philanthropy for the high-MEDIA subsample and for the low-MEDIA subsample, respectively. Also, untabulated results suggest that CER_WEAK is significantly positively associated with corporate philanthropy for the high-MEDIA subsample (.4160 with t = 3.52), but the coefficient on CER weakness is negative and insignificant for low-MEDIA subsample (−.0656 with t = −.44).⁶ These asymmetric findings suggest that high media coverage increases the likelihood of the exposure of CER weakness (environmental wrongdoing, misconduct, or even greenwashing), and thus a firm is inclined to engage in corporate philanthropy to mitigate or offset the negative influence of CER weakness. However, low media coverage does not motivate a firm with CER weakness to carry out corporate philanthropy. Moreover, above asymmetric findings can borrow support from Delmas and Cuerel Burbano (2011) and Du (2015a) that attach the importance of disclosing greenwashing or CER weakness to media coverage.

Fig. 1

The interaction between CER weakness (CER_WEAK) and media coverage (MEDIA) on corporate philanthropy (GIVING). Note: In Fig. 1, black, red, and blue lines denote the influence of CER weakness on corporate philanthropy for the full sample, the low-MEDIA subsample, and the high-MEDIA subsample, respectively

Robustness checks of Hypotheses 1 and 2

Table 3 provides robustness checks using alternative measures of corporate philanthropy and corporate environmental responsibility weakness.

Table 3 Robustness checks using alternative variables of CER weakness and corporate philanthropy Panel A: Robustness checks of Hypotheses 1 and 2 using the raw CER weakness

First, we use the raw CER weakness (CER_WEAK_RAW), measured as “Max_ENV-ENV _{j, t}” (Max_ENV denotes the maximum value of ENV _{j, t} by industry and year; ENV _{j, t} denotes corporate environmental performance for firm j in year t), as the main independent variable to re-test Hypotheses 1 and 2. As shown in Column (1) of Panel A, the coefficient on CER_WEAK_RAW is significantly positive (.0089 with t = 2.89), providing additional and strong support to Hypothesis 1. As Column (3) shows, the coefficient on CER_WEAK_RAW×MEDIA is positive and significant at the 1 % level (.0067 with t = 2.66), lending important support to Hypothesis 2. Moreover, both CER_WEAK_RAW and MEDIA have significantly positive coefficients.

Second, we use the rank of CER weakness (CER_WEAK_RANK) as the independent variable. As Column (1) of Panel B shows, CER_WEAK_RANK has a significantly positive coefficient (.0005 with t = 2.84), additionally supporting Hypothesis 1. In Column (3) of Panel B, the coefficient on CER_WEAK_RANK×MEDIA is positive and significant at the 5 % level (.0003 with t = 2.40), consistent with Hypothesis 2. Furthermore, the coefficients on both CER_WEAK_RANK and MEDIA are significantly positive, consistent with Hypothesis 1 and findings in Table 2.

Finally, we use the rank of corporate philanthropy (GIVING_RANK) as the dependent variable. GIVING_RANK is the ordered variable for corporate philanthropy, measured as 10, 9, 8…… and 1 for the first 10 %, the second 10 %,......, and the tenth 10 % of corporate philanthropy, respectively, and 0 for firms without corporate philanthropy. As shown in Column (2) of Panel C, the coefficient on CER_WEAK is significantly positive (1.5412 with t = 1.86), consistent with Hypothesis 1. In Column (4) of Panel C, we find that the coefficient on CER_WEAK × MEDIA is positive and significant at the 1 % level (2.3427 with t = 2.97), lending additional and strong support to Hypothesis 2.

Overall, results in Table 3 suggest that our main conclusions are not qualitatively changed using alternative measures of corporate philanthropy and CER weakness.

The potential endogeneity issue

Theoretically, we cannot rule out ex ante the potential endogeneity between CER weakness and corporate philanthropy. On the one hand, firms with CER weakness may use corporate philanthropy to alleviate the negative influence of their environmentally unfriendly behavior. On the other hand, “firms might risk greater environmental misconduct if they believe that their philanthropy will help them get relatively small punishments” (Du, 2015b: 356). Thus, we use two-stage OLS-Tobit regression procedures to mitigate the endogeneity between CER weakness and corporate philanthropy in the simultaneous setting. Specifically, we employ Model (5) in Table 4 to predict the expected CER weakness (CER_WEAK*), and then we estimate Models (2) and (4) after including the predicted CER_WEAK*.

In Model (5), ROA is return on total assets, measured as operating income divided by total assets. FIN is the amount of equity capital or debt raised during the year divided by total assets (Clarkson et al., 2008). VOLAT denotes stock price volatility, measured as standard deviation of market adjusted weekly stock return (Clarkson et al., 2008). CAPIN denotes capital intensity, measured as the ratio of capital spending (including fixed assets, intangible assets and other long-term assets) scaled by total sales revenue (Clarkson et al., 2008). LISTAGE is the number of years since a firm’s IPO. In fact, in this study, ROA, FIN, VOLAT, CAPIN, and LISTAGE can serve as five instrument variables.⁷

Next, we use CER_WEAK* as the independent variable to estimate Model (2) and Model (4) and test Hypotheses 1 and 2. Table 4 reports results using two-stage OLS-Tobit regressions. In particular, Column (1) and Model (5) present results of the first-stage regression, and we report the second-stage regression results of Hypotheses 1 and 2 in Columns (2)–(3) (Models (2) and (4)), respectively.

Table 4 Robustness checks of Hypotheses 1 and 2 considering the potential endogeneity between CER weakness and corporate philanthropy

In Column (1) and Model (5) we find that CER weakness is significantly positively associated with ROA, CAPIN, LISTAGE, BIG10, and INDR. Moreover, CER weakness is significantly negatively related with FIRST, ANALYST, SIZE, OCF, INTER, IND_REG, and TOBIN’Q. ⁸

In Column (2) and Model (2), the coefficient on CER_WEAK* is positive and significant at the 5 % level (1.9690 with t = 2.21), providing strong support to Hypothesis 1. In Column (3) and Model (4), we find that the coefficient on CER_WEAK*×MEDIA is positive and significant at the 1 % level (.3771 with t = 3.70), lending important support to Hypothesis 2. Furthermore, both CER_WEAK* and MEDIA have significantly positive coefficients, consistent with findings in Table 2.

In summary, results using two-stage OLS-Tobit regressions reinforce our findings in Table 2.

Discussions

Theoretical contributions

This study contributes to the existing literature in several ways. First, to our knowledge and literature in hand, this study is one of very few studies, if it is not the first one, to examine the hidden connection between CER weakness and corporate philanthropy. Previous literature rarely examines directly the relationships among different CSR dimensions such as environmental responsibility, corporate philanthropy, employee relations, and product safety. Our study documents the association between CER weakness and corporate philanthropy and thus lends important support to the view that various CSR dimensions may be not inherently consistent. In addition, this finding also implies that Chinese enterprises may carry out corporate philanthropy as the window-dressing of environmental wrongdoing, echoing and validating the argument in Delmas and Cuerel Burbano (2011).

Our study also distinguishes itself from Du (2015b). Using the amount of fines imposed by regulators as the proxy for environmental misconduct, Du (2015b) based his findings on extreme cases or environmental violations (environmentally illegal activities), and thus conclusions may not fit in well with the association between CER weakness (a relatively continuous variable) and corporate philanthropy. It is well-known that polluting industries are controversial and the public continues to torture the conscience of polluting enterprises. In addition, firms in polluting industries also try every shift available to divert stakeholders’ attention away from the negative events and inverse images caused by CER weakness. As a result, comparatively, our study provides more direct evidence on the hidden association between CER weakness and corporate philanthropy. Moreover, Du (2015b) only employed the survey data of 2008. Considering the limitation of the survey data, Du (2015b) only controlled several internal governance mechanisms and financial characteristics, but did not include other important determinants such as external monitoring mechanisms and conditional factors. Therefore, Du (2015b) did not base his study on the comprehensive framework of factors influencing corporate philanthropy. As a result, our findings are more robust, compared with Du (2015b).

Second, this study explores and recognizes CER weakness dressing as an additional motivation of corporate philanthropy. Borrowing the support from extant literature (Zyglidopoulos et al., 2012), we use the classification between CSR-strength and CSR-weakness to discuss the relationship between CER weakness and corporate philanthropy. Extant studies have recognized strategic, altruistic, political, and managerial self-interested motivations for corporate philanthropy (Brammer & Millington, 2005, 2006; Campbell, Moore, & Metzger, 2002; Du, 2015b; Saiia et al., 2003; Seifert, Morris, & Bartkus, 2003). Our study documents the positive association between CER weakness and corporate philanthropy, providing some important supplements to extant studies about the motivations of corporate philanthropy.

Third, this study echoes and adds to the existing literature on media coverage’s influence on CSR (Zyglidopoulos et al., 2012). We find that media coverage is significantly positively associated with corporate philanthropy, and further media coverage reinforces the positive association between CER weakness and corporate philanthropy. These findings attach the importance to the governance role of media coverage in inspiring corporate philanthropy and further the reinforced effects of media coverage on the positive association between CER weakness and corporate philanthropy.

Fourth, using the context of China, our findings contribute to previous literature based on the context of developed markets in which scholars can observe more pronounced causality between CER weakness and corporate philanthropy. Firms in developed markets are confronted with enormous pressure to carry out corporate philanthropy and engage in environmental conservation (Du, 2015b; Sharfman & Fernando, 2008), but in emerging markets where the existing laws and regulations work less effectively and business ethics are being formed (Du et al., 2014), many enterprises lack the consciousness of philanthropy and environmental conservation (Du, 2015b; Sharfman & Fernando, 2008). Therefore, it remains uncertain whether findings in extant studies based on developed markets can fit in well with the Chinese context. In this regard, our findings provide supplementary evidence to those based on developed markets.

Finally, following GRI (2006) and Clarkson et al. (2008), we calculate CER weakness based on the multi-dimensional evaluation index system for corporate environmental performance, including seven components and 44 sub-components. The measurement has the advantage over other measures (e.g., fines due to environmental pollution, environmental expenditures, etc.) in extant literature that only emphasize one or several aspects of corporate environmental responsibility, ensuring that our study better captures a firm’s efforts in environmental conservation. In this regard, we can further investigate the causality between CER weakness and corporate philanthropy.

Managerial implications

Our study has several managerial implications. First, we validate the positive association between CER weakness and corporate philanthropy, implying that environmental wrongdoing-dressing is an important motivation of corporate philanthropy. This finding supplements the existing four motivations of corporate philanthropy (i.e., strategic, political, altruistic, and managerial self-interested motivations), which calls attention of investors to a firm’s real motivation or intention when it engages in corporate philanthropy. Specifically, stakeholders should not attribute a firm’s engaging in corporate philanthropy for the purpose of covering its environmental wrongdoings to altruistic motivation. Moreover, our study inspires the public, practitioners, and stakeholders to pay close attention to a firm’s greenwashing behavior, including the motivations of greenwashing, the conduits of greenwashing, and the economic consequences of greenwashing.

Second, this study finds that media coverage is significantly positively associated with corporate philanthropy and media coverage reinforces the positive association between CER weakness and corporate philanthropy. These findings reveal that media coverage, as an important informal system, can play its roles of information intermediary and monitoring, which is very important for emerging markets like China where formal institutions are incomplete or performed less effectively and business ethics are under construction (Du et al., 2014). In this regard, media coverage can serve as an alternative to weak formal institutions and thus motivates Chinese enterprises to carry out corporate philanthropy in substance, rather than as a channel to mitigate the negative influence of their environmental wrongdoings or CER weaknesses.

Finally, our findings can be applied to other emerging markets where environmental deterioration becomes increasingly alarming. In fact, the real motivations of corporate philanthropy in developing markets are complex and compound, and it is not as simple as altruistic or ethical behavior. More specifically, our findings can help regulators, stakeholders, and the public to exempt from being fooled by pseudo-philanthropy that may be proved to be window-dressing of other corporate wrongdoings.

Limitation and future research

Although this study examines the association between CER weakness and corporate philanthropy, it has several limitations that may be addressed in future research. First, because of data limitations, we focus mainly on the inherent inconsistency between corporate philanthropy and environmental responsibility, and thus future research can focus on other specific CSR dimensions and examine whether wrongdoing-dressing exists among other CSR dimensions (e.g., product safety, employee relations, etc.) and corporate philanthropy. Moreover, future research can focus on how enterprises in polluting industries conduct greenwashing and how the market values corporate greenwashing behavior, which can add to the existing literature on greenwashing (Delmas & Cuerel Burbano, 2011). Second, maybe it is more important for researchers to theoretically optimize the framework of factors influencing corporate philanthropy and the motivations of corporate philanthropy, combined with systematic evidence on the one hand, and to identify different channels in which firms conduct greenwashing on the other hand. Third, future research should develop more comprehensive methodologies to measure corporate environmental performance, which ensure researchers to conduct their studies on the basis of comparability. Finally, this study is conducted in the Chinese context, and thus findings in this study may not be directly applied to other developed markets. Future research should extend to different markets and use the international setting to examine the potentially conflicted relationships among different CSR dimensions.

Conclusion

Using a sample of Chinese listed firms in polluting industries, our findings show that CER weakness is significantly positively associated with corporate philanthropy, suggesting that firms with CER weakness use corporate philanthropy as an important channel to mitigate the negative influence of environmentally unfriendly behavior on their corporate images. Moreover, media coverage reinforces the positive association between CER weakness and corporate philanthropy, validating the monitoring role of media coverage in emerging markets like China. In a nutshell, our study uncovers a typical greenwashing behavior in which many Chinese enterprises in polluting industries try to offset the negative impacts of their environmental weaknesses (wrongdoings) by carrying out corporate philanthropy.

Appendix

Table 5 The procedures and descriptive statistics of corporate environmental performance (raw) score