University of Birmingham Do foreign institutional investors drive corporate social responsibility? Evidence from listed firms in China

This paper investigates the effect of qualified foreign institutional investors (QFIIs) on corporate social responsibility (CSR) within the context of listed firms in China. We find that QFIIs offer an incisive channel for improving socially responsible practices. In addition, we find that firms with QFIIs are more likely to comply with the Global Reporting Initiative (GRI) guidelines, and that their sustainability reports tend to be longer. We also find that this positive effect is more pronounced in firms with low initial CSR scores than those with high CSR scores at the time when QFIIs enter the sample. Our empirical evidence further confirms that this positive impact is driven by QFIIs from countries with high social awareness, or QFIIs from geographically distant countries, consistent with their motives, and is linked to the ownership of QFIIs,especially when theQFIIis among thetoptenofthe largest shareholders. Finally, our extended analysis reveals that the increase in CSR performance associated with the presence of QFIIs results in greater firm performance and easier access to finance. literature on the role of foreign institutional investors. Prior studies exploring the effects of foreign ownership mainly focus on corporate governance (Huang & Zhu, 2015), firm performance al., 2006), accounting quality et al., 2017), and the integration of human and financial resources (Huang & Shiu, 2009). This study enriches the literature on the stakeholder-oriented implications of foreign ownership and highlights the importance of foreign investors in deploying social-engagement activities in the Chinese market.

for firms to engage in better CSR, the social and environmental issues are still prevalent, mainly due to weak law enforcement (Elmagrhi et al., 2019). Second, the Qualified Foreign Institutional Investor (QFII) Scheme launched by the government in 2002 has attracted a large number of foreign institutional investors, especially those from developed markets, to listed firms in China. The QFII Scheme allows foreign owners to gain access to the domestic capital market, and these offshore investors are expected to play an essential role in corporate decisions and governance practices. For example, Huang and Zhu (2015) state that the role of QFIIs in promoting Chinese governance is significant with regard to split-share structure reform. Prior studies show that the foreign parties investing in emerging markets can potentially enhance the information environment in association with an increase in analyst coverage and a decline in earnings management, resulting in enhanced firm innovation (Lel, 2019;Luong et al., 2017). Given the important role of QFIIs in China, assessing the extent to which foreign investors influence social responsibility provides new and critical insights that may link changes in business strategy with the push towards effective CSR engagement and commitment.
McGuinness, Vieito, and  explore the role of foreign ownership, as one of their research questions, on social ratings in China, and report the absence of a significant positive impact of foreign investors on CSR performance. 3 With an extended sample period, our study builds upon and goes beyond McGuinness et al. (2017) in order to fill this important research gap in the CSR literature. Based on a sample of 752 listed firms with 4,145 firm-year observations and QFIIs covering 22 countries during the period 2009-17, we provide evidence that the presence of foreign institutional investors has a positive effect on CSR policies, and this positive impact is economically significant, suggesting that CSR performance, on average, is approximately 4.28% higher for firms with QFIIs than those without QFIIs, ceteris paribus. In addition, our results reveal that firms with foreign institutional investors are more likely to comply with the Global Reporting Initiative (GRI) guidelines, and their reports on sustainability tend to be longer and more comprehensive. Additionally, we find that such a positive effect is more pronounced in firms with low initial CSR performance than those with high initial CSR performance, suggesting that the scope of the impact depends on initial CSR performance at the time when QFIIs enter the sample. Our finding is robust to several additional tests, including the Granger causality test, a firm fixed-effect model, and a weighted least squares approach, and controlling for additional effects of cross-listing, R&D intensity, excess cash, and a sub-sample excluding firms in the manufacturing sector.
More importantly, our result holds after endogeneity concerns are carefully addressed, based on alternative research designs, such as the propensity score matching (PSM) and the Heckman two-stage selection analysis.
Next, we investigate why QFIIs push firms towards better CSR practices. The incentive for QFIIs to promote CSR initiatives may be attributable to their high level of social awareness due to the fact that approximately 90% of foreign investors in Chinese listed firms are from developed economies or well-governed regimes, such as countries in North America and Western and Northern Europe, where CSR engagement is seen as desirable. When investing in foreign markets, these offshore owners are highly likely to transplant their social awareness to the firms that they invest in and pay particular attention to environmental and social issues. In addition, investing in a foreign country is associated with a unique risk due to the nature of these investments in foreign markets, such as a lack of adequate and transparent information for rationally assessing their target firms, and foreign investors are often assumed to have an information disadvantage relative to domestic investors (Barmeyer & Mayrhofer, 2008;Gehrig, 1993). Under such circumstances, promoting socially responsible activities may thus be considered as a way to address such concerns. Indeed, CSR engagement is generally viewed as an important signalling mechanism that reduces the information asymmetry (Dhaliwal, Li, Tsang, & Yang, 2014;Dhaliwal, Li, Tsang, & Yang, 2011). Consistently, our results reveal that this positive impact stems from two channels. In particular, we find that this positive impact is more pronounced in firms with foreign investors from countries with high regulatory quality than those from countries with low regulatory quality, and in firms with foreign investors from geographically remote countries than those from physically proximate nations. 3 Based on the sample covering the period 2009-13, McGuinness, Vieito, and indicate the absence of significant positive effect of foreign parties, including both QFIIs and Overseas, on social ratings (see p. 86). In particular, their results presented in tables 3-5 show that the impact of QFIIs on CSR is negative, although marginally significant. We next pin down the mechanism for such a positive impact. The literature suggests that large and key shareholders can exert influence on corporate policies through proposing and voting, or by appointing directors on the board or shareholder activism (Lee & Lounsbury, 2011;Oh, Chang, & Martynov, 2011). Using percentage of ownership by QFIIs to capture the degree of influence, and whether a QFII is among the top ten shareholders as a measure for the strength of influence, we provide evidence suggesting that CSR performance is positively linked to the ownership of QFIIs, and this positive link is more salient if the QFII is among the top ten shareholders, suggesting that the mechanism through which QFIIs influence CSR is from the voting power, especially by appointing a director on the management team. Overall, our results convey that foreign investors, when their investment stakes are significant, have a greater influential scope on corporate decision making and offer a key channel that links the changing business strategy that pushes towards effective CSR engagement and commitment. Finally, our extended study reveals that the enhanced CSR performance driven by the presence of foreign institutional investors leads to real consequences on business strategies in terms of greater operating performance and easier access to finance. This paper contributes to the literature in several ways. First, this study adds insights to the general literature on the determinants of CSR. Prior research has documented that CSR strategies are influenced by academic directors (Cho, Jung, Kwak, Lee, & Yoo, 2017), cross-listing (Boubakri et al., 2016), CEO marital status (Hegde & Mishra, 2019), regulatory and legal systems , and board gender diversity (Bear, Rahman, & Post, 2010;Elmagrhi et al., 2019;McGuinness et al., 2017). Our study highlights the importance of foreign investors, especially their social awareness and geographical distance, as a critical channel to promote better social practices for firms listed in countries, such as China, with weak institutional environments.
Second, this paper provides new evidence to the ever-growing literature on the role of foreign institutional investors. Prior studies exploring the effects of foreign ownership mainly focus on corporate governance (Huang & Zhu, 2015), firm performance (Douma et al., 2006), accounting quality (Beuselinck et al., 2017), and the integration of human and financial resources (Huang & Shiu, 2009). This study enriches the literature on the stakeholder-oriented implications of foreign ownership and highlights the importance of foreign investors in deploying social-engagement activities in the Chinese market.
The paper proceeds as follows. In Section 2 we review the literature and develop hypotheses, with the research design presented in Section 3. Section 4 includes a discussion of the main findings, robustness checks, and endogeneity tests. In Sections 5 and 6 we explore the motives for QFIIs in adopting better CSR practices and reveal the mechanism for such an impact. Section 7 provides further analysis, and Section 8 concludes.

Background
The QFII scheme launched by the authorities in 2002 is one of the most important open-door policies in the Chinese equity market. Based on this scheme, foreign institutional investors are allowed to invest directly in the RMBdenominated capital market. The motive of introducing foreign investors to the Chinese domestic market is generally considered twofold. First, the Chinese market, as a relative latecomer to global capital markets, attempts to rely on foreign institutional investors to effectively integrate human capital and financial resources, in addition to introducing superior management skills and governance practices. Second, the QFII scheme is expected to suppress the impact of overseas speculative 'hot money' on the domestic economy, in building up an open, competitive, and orderly modern market system (Huang & Zhu, 2015). Since the launch of this scheme, foreign investors in domestic A-share markets have increased significantly in terms of investment quota and number of QFIIs, from only one QFII with an initial quota of $800 million (made by UBS AG, the Swiss investment bank), in July 2003, to 283 QFIIs with a total investment quota of $93 billion, as of 31 July 2017. 4 The recent surge in the number of foreign investors in the Chinese domestic market is likely to exert direct influence on firms' corporate governance, management culture, and monitoring. For example, a recent paper by Huang and Zhu (2015) asserts that QFIIs promote minority shareholders' interest in the split-share structure reform process. The underlying premise is that the existence of foreign investors enhances corporate governance in China. Li, Brockman, and Zurbruegg (2015) provide evidence that cross-listed H-shares traded by foreign investors incorporate significantly more firm-specific information than their A-share counterparts traded by domestic Chinese investors, thereby generating cross-listing value benefits and confirming the positive effects of QFIIs on the governance system.
In the literature, however, how corporate social activities are influenced by the presence of QFIIs is largely unexplored. China has recently promoted CSR practices, driven by high demand for corporate social awareness in addressing environmental and social issues, such as air pollution and food safety. Specifically, China, whose coal energy makes up approximately 70% of total global energy consumption, has become the largest carbon-dioxide polluter in the world in recent years (Elmagrhi et al., 2019). Further, the Chinese milk scandal in 2008 that affected more than 300,000 children was referred to by the World Health Organisation as one of the most significant food safety incidents in recent years. Indeed, China has made substantial progress in embracing CSR in the past few years by introducing a series of localised sustainability reporting guidelines. 5 In addition, both the Shenzhen and Shanghai Stock Exchanges enacted the Guidelines on Social Responsibility and Environmental Information Disclosure in 2006 and 2008, respectively, to strengthen social consciousness and help listed firms achieve broader social objectives. Despite all these efforts, the environmental protection and social welfare systems, as well as the rules for labour and human rights protection, are all required to be in accordance with international standards. Thus, we believe that the current study is timely and important, as the opening of the domestic market to foreign investors, especially those from developed markets, is likely to have a significant influence on the corporate culture, motivating firms to further embrace CSR activities.

Related literature and hypothesis development
Corporate social responsibility is broadly defined in the literature as a firm's commitment to minimising potential harmful effects of its operations on its stakeholders and maximising its long-run beneficial impact on society (Boubakri et al., 2016;Ferrell et al., 2016;Kim, Li, & Li, 2014;. Although there is a considerable debate regarding the economic and social desirability of CSR, the recent literature largely focuses on the positive view of CSR. 6 Drawing on stakeholder theory, the positive view of CSR suggests that firms rely on various stakeholders, such as owners, employees, consumers, and the community as well as the society at large, to support the business in achieving long-term strategies, such as reputation, legitimacy, and superior firm performance (Cheung, Kong, Tan, & Wang, 2015;Ferrell et al., 2016). In addition, Siegel and Vitaliano (2007) contend that socially responsible activities may provide firms with an avenue to differentiate themselves from other firms and signal their ethical behaviours and trustworthiness to the capital market, which, in turn, helps firms gain capital market benefits (Deng et al., 2013;Ferrell et al., 2016). In particular, socially responsible firms are found to be more likely to constrain earnings management and to make responsible operating decisions, thereby maintaining transparency in financial reporting and reducing information asymmetries (Kim, Park, & Wier, 2012 (Dhaliwal, Li, Tsang, & Yang, 2014;Dhaliwal, Li, Tsang, & Yang, 2011); build the reputation for firms (Brammer & Millington, 2005); reduce firm financial risk (Benlemlih & Girerd-Potin, 2017); and enhance the earnings quality (Kim, Park, & Wier, 2012), investment efficiency and innovation (Benlemlih & Bitar, 2018;Cook, Romi, Sánchez, & Sánchez, 2019), and financial performance ( should have a strong incentive to promote socially responsible activities, which are inevitably beneficial for business in the long run. Compared to the domestic investors whose demands for CSR are motivated by financial returns, foreign investors' requests for strong CSR activities may reflect two additional aspects. First, CSR practices have been well adopted for several decades in well-governed economies. When investing in markets with weakly governed institutional environments, such as China, these offshore owners may show similar behaviours and exert direct influence on the management team; for example, transplanting their high social awareness to the firms that they invest in. Second, investing in a foreign country is associated with additional risk due to cultural differences and geographical distances, and foreign investors are often assumed to have an information disadvantage relative to domestic investors (Barmeyer & Mayrhofer, 2008;Gehrig, 1993). Under such circumstances, foreign parties may pay more attention to social and environmental issues because the role of CSR engagement can be viewed as an essential signalling mechanism in reducing information asymmetry (Oh et al., 2011). In sum, foreign investors have great incentives in promoting the social conduct of listed firms because by doing so, they can share the financial success of the entities they invest in.
Empirical evidence relating to the effect of foreign investors on CSR is limited. Cheung et al. (2015) report that there is a demand for internationalised firms in China to comply with CSR practices, driven by pressure from foreign stakeholders, in order to maintain contracts and sustain business relationships. Oh et al. (2011) find that the percentage of foreign ownership is positively associated with CSR performance in the Korean market, consistent with the view of Chapple and Moon (2005) who document that globalisation enhances firms' CSR engagement in Asian countries.
Therefore, our first hypothesis is drawn from the discussion above, and is stated formally below: H1: Firms with the presence of QFIIs exhibit higher CSR performance than those without QFIIs.

Compared with markets characterised by better social engagement and corporate governance, CSR involvement in
China is still at the initial stage in terms of wide adoption of regulatory guidelines, codes of social conduct and regulations (Husted, 2015;Sethi, Martell, & Demir, 2017). Although environmental conditions and social issues have forced the government to develop a series of regulations, there is a significant disparity in the adoption of CSR engagement among the listed firms in China. For example, the range of overall CSR scores in our sample is from 14.14 to 87.95.
When measured at the industry-adjusted level, such divergence is even more pronounced, varying from −24.31 to 49.27, and more than half of the firms in our sample have CSR scores lower than the average industry level. This disparity leads us to contemplate that the scope of the influence from QFIIs may depend on the firm's initial CSR engagement at the time when foreign investors place their investments. If foreign institutional investors indeed advocate for meaningful CSR practices, we should perceive a greater effect in firms with lower CSR performance than those with higher CSR performance at the time when foreign owners entered the sample. Thus, our second hypothesis is formally stated below:

H2:
The impact of QFIIs on CSR performance is more pronounced in firms with low CSR performance than those with high CSR performance at the time when QFIIs enter the sample.

Sample construction and data sources
The firm-level sample includes all firms incorporated in China and publicly traded on either the Shanghai Stock Exchange or the Shenzhen Stock Exchange with environmental and social ratings between 2009 and 2017. The proxies for environmental and social performance are obtained from the RKS rating system, 7 which details the contents of CSR-related activities and fully covers Chinese listed firms issuing CSR or sustainability reports (Elmagrhi et al., 2019;Lau, Lu, & Liang, 2016;Luo, Wang, & Zhang, 2017;Marquis & Qian, 2014;McGuinness et al., 2017). The CSR performance score ranges from 0 to 100, with higher values corresponding to a better quality of social conduct.
Information on the ownership identities and characteristics-for example, foreign institutional ownership, shareholdings, and name and headquarters of each foreign institution on listed firms in China-is obtained from the

Key variables
The aggregate CSR performance (CSR), a measure of the overall environmental and social performance of a firm in a given year, represents the quality, effectiveness, and content of Chinese listed firms' overall CSR activities. The overall CSR performance consists of three dimensions, defined under the headings of Macrocosm, Content, and Technique.
Macrocosm performance evaluates the overall strategy and disclosure channels employed in a firm's CSR reporting.
Content, the most important dimension, measures a firm's effectiveness of the implemented environmental strategy, economic performance, labour and human rights protection, and community participation. Technique measures the extent, coverage, and consistency of social reporting. 8 In addition to raw CSR ratings, we adopt several alternative measurements as proxies for the quality of socially responsible activities. First, we consider whether a firm adopts the GRI guidelines, which is measured as a dummy variable with a value equal to 1 if a firm voluntarily discloses sustainability or CSR reports following the GRI reporting standards, and 0 otherwise. The GRI guidelines are viewed by communities and regulatory authorities all over the world as a way to communicate firms' impact on important environmental and social issues such as environmental protection, resource depletion, carbon emissions, climate change, and social well-being (Marquis & Qian, 2014).
Therefore, compliance with the widely accepted GRI standards is generally recognised as acting environmentally and socially responsible, as the GRI guidelines are deeply rooted in the public interest (Levy, Szejnwald Brown, & De Jong, 2010).
Next, we employ the length of environmental and social reports (Page), defined as the number of pages of the firms' CSR report in natural logarithm, as an additional measurement of CSR. In disclosing overall CSR ratings, some firms choose to state such information in their annual financial reports or sustainability reports, and others are more likely to have stand-alone CSR reports with better readability and more reporting contents in order to demonstrate their special commitment and efforts to improve transparency regarding the risk management process and performance in the long run. Dhaliwal et al. (2011) contend that investors require a lower rate of return for firms whose stand-alone CSR or sustainability reports are more comprehensive and detailed. Finally, we utilise industry-adjusted CSR performance to evaluate firms' environmentally and socially friendly investments relative to their peers in the same sector, as CSR scores vary considerably across industries. The industry-adjusted CSR (CSR_IA) is measured as the deduction of the overall CSR rating of a firm from the mean rating for all listed firms in the same industry for a given year. 9 Qualified foreign institutional investors, indicated by QFII, is the variable of interest in this study, and is measured as a dummy variable with a value equal to 1 if a firm has foreign institutional investors, and 0 otherwise. 7 The RKS rating system, developed by an independent and leading rating agency in China, follows the evaluating system of the MSCI ESG STATS database (formerly known as KLD) and standards of GRI activity (GRI3.0) to construct its own rating system. It provides a comprehensive measure that reflects the CSR practices of a firm. More information is available at http://www.rksratings.cn/.

Model specification
To investigate the relationship between QFIIs and CSR performance, we specify the following regression model with fixed effects following Boubakri et al. (2016), McGuinness et al. (2017), and Dyck et al. (2019): where CSR represents the CSR performance rating; QFII stands for qualified foreign institutional investors; and Control is a vector of variables that may have potential effects on CSR performance. Specifically, the variables used to control for the effects of firm-level characteristics are state ownership (State), firm size (Size), firm age ( The variable construction is shown in Appendix B.   Table 3 presents correlations of the main variables. It is evident that the correlation coefficient between QFII and CSR is positive and significant. This lends support to the initial finding of the positive influence of foreign investors on CSR performance. None of the other variables are highly correlated, ruling out potential multicollinearity. Mining (B) 37

TA B L E 1 Distribution of foreign institutional investors
Manufacturing (C) 567 Electricity, gas and water production, and supply (D) 41 Construction ( Notes: We obtain the country-level regulatory quality scores from the World Bank and calculate the median value for each year. If a QFII is domiciled in a country with a regulatory quality score equal to or above the median value in a given year, it is classified as high regulatory quality, otherwise, low regulatory quality. Similarly, we calculate the median value of the geographical distances between the capital cities of QFIIs' countries of domicile and the capital city of China, and divide QFIIs into those from remote-distance countries (equal to or above-median physical distance) and near-distance countries (belowmedian physical distance).

Variable
No. of Obs.

Main regression results
To explore the impact of foreign institutional investors on CSR, we estimate equation (1) and report the results in Panel A of Table 4. Model 1 is our baseline model, where the dependent variable is CSR score. Notably, the coefficient on QFII is positive and statistically significant (t = 3.26), suggesting that firms with QFIIs experience an increase in CSR performance relative to those without QFIIs. 10 Importantly, this positive effect is economically significant, indicating that CSR performance is, on average, approximately 4.28% higher for firms with QFIIs than those without, ceteris paribus. 11 Therefore, the evidence supports hypothesis H1.
The effects of control variables on CSR performance, if significant, are all broadly consistent with previous evidence in the literature. Specifically, CSR performance is positively associated with firm size, Tobin's Q, free cash flow, ownership concentration, size of board, and managerial size, and negatively related to firm age, leverage, return on assets, CEO-Chairman duality, and SSE. It is worth noting that the coefficient on State is not significant, suggesting that the government is not a significant factor in promoting CSR performance. Our study is consistent with Cheung et al. (2015), who suggest that state-owned enterprises are not particularly different from non-state-owned firms in responding to pressure on CSR activities.
As to other alternative measures for a firm's environmentally and socially friendly behaviour, the dependent variable is GRI in Model 2, estimated using a panel probit model, and Page in Model 3. In both specifications, the coefficient on QFII is positive and statistically significant, suggesting that firms with QFIIs are more likely to issue sustainability reports following the GRI guidelines, and have a potential effect on the details and contents of environmental and social reporting relative to firms without QFIIs. Finally, using the industry-adjusted CSR performance (CSR_IA) as an explanatory variable shown in Model 4, it further confirms the positive impact of QFIIs on CSR performance. Overall, using a wide range of alternative measures of CSR, the evidences reaffirm that the presence of QFIIs pushes firms to adopt better CSR engagement.
Note that in all model specifications shown in Table 4, Age is negative and statistically significant. 12 The result seems to be in contrast to the argument that young firms may not have sufficient resources to invest in CSR or environmental protection activities, predicting a positive link between firm age and CSR exists (Boubakri et al., 2016;Withisuphakorn & Jiraporn, 2016). However, in a highly competitive environment as seen in China, young or start-up firms may have an incentive to actively engage in CSR as an instrument to establish a good reputation, consistent with the perspectives of Brammer and Pavelin (2006), in order to compete with mature firms for bank loans and market participants. For example, Wens Foodstuff Group (stock code: 300498), the firm listed on the ChiNext of Shenzhen Stock Exchange in 2015, is a relatively young firm, but its total social welfare donations and environmental protection investments amounted to approximately 100.55 and 157.10 million USD in 2016 and 2017, respectively. 13 Next, to further explore the extent to which QFIIs influence CSR performance, we divide the sample into high and low CSR firms at the time when QFIIs make their investments. In doing so, we first calculate the median value of CSR performance for each year and then, based on this, categorise a firm into a higher CSR group if its initial CSR score in a given year is above or equal to the median CSR level, and a lower CSR group, if otherwise. We then re-run equation (1) separately for each subsample and report the results in Table 4, Panel B. It shows that the coefficient on QFII is positive for both groups, and it is highly significant in firms with low initial CSR. Consistent with hypothesis H2, our result indicates that the effect of QFIIs is more pronounced in firms with low initial CSR performance than those with high 10 We also use CSR scores in natural logarithm as an alternative measurement, and the result is unaffected. In addition, we investigate the impact of QFIIs on CSR dimensions, and our results show that firms with QFIIs promote all three categories including Macrocosm, Content, and Technique performance. All results are available upon request.  (1), the compliance with the Global Reporting Initiative (GRI) in Model (2), the natural logarithm of the length of CSR reports (Page) in Model (3), and the industry-yearadjusted CSR performance (CSR_IA) in Model (4). In addition, marginal effects are reported in Model (2). Panel B reports the results by dividing the sample into high and low CSR firms at the time QFIIs enter the sample. Variable definitions can be found in Appendix B. In all specifications, independent variables are lagged by one year. Standard errors are clustered at the firm and year level and are corrected for heteroscedasticity. All regressions control for industry and year fixed effects. T-statistics (or Z-statistics) are reported in parentheses. *** , ** , and * indicate statistical significance at the 1%, 5%, and 10% level, respectively.
initial CSR. In other words, the influence of foreign institutional investors is more salient in firms with more scope for CSR improvement. Following Henderson, Kumbhakar, Li, and Parmeter (2015) and Dyck et al. (2019), we find that the difference between two subsamples is statistically significant (p-value = 0.0363) based on the seemingly unrelated estimation.

Robustness tests
In this section, we perform several robustness tests to examine the validity of our main finding by employing the Granger causality test, a firm fixed-effect approach, and weighted least squares, as well as controlling for additional effects of cross-listing, R&D intensity, and excess cash, and using a sub-sample excluding the firms in the manufacturing sector. We explain our results based on the baseline model for brevity.

Granger causality test
Reverse causality may potentially bias our results because foreign investors may choose to invest in socially responsible firms. For example, Wang and Chen (2017) find that foreign institutional investors tilt their investment in firms that are doing well at saving energy. To rule out reverse causality as an alternative explanation for our results, we turn to the Granger causality test by estimating two symmetric sets of regressions (Dyck et al., 2019); that is, we regress CSR on lagged QFII and lagged CSR, and QFII on lagged CSR and lagged QFII, with the same set of lagged control variables, separately. The results, displayed in Models 1 and 2 of  (1)-(2), a firm fixed-effect model in Model (3), weighted least squares (WLS) in Model (4), the inclusion of Cross, RDIntensity and ExcessCash in Models (5) to (7), respectively, and a sub-sample without manufacturing sector in Model (8). In the firm fixed-effect model, SSE is automatically excluded because there is no change in a firm's exchange-listed status over any given two years. All variables are defined in Appendix B. In all specifications, independent variables are lagged by one year. Standard errors are clustered at the firm level and are corrected for heteroscedasticity. T-statistics (or Z-statistics) are reported in parentheses. *** , ** , and * indicate statistical significance at the 1%, 5%, and 10% level, respectively. statistically significant, but that on CSR is not. 14 Thus, we conclude that it is the QFIIs that drive firms' social performance, rather than the firms with high CSR performance that attract the investments of foreign institutional investors.

Firm fixed-effect approach
In the previous analysis, we control only for observable firm characteristics. Some unobservable factors may, however, drive the results. To mitigate potential concerns that our results may arise from omitting time-invariant firm-specific factors, we follow Hegde and Mishra (2019) to estimate equation (1) by employing a firm fixed-effect approach. The result is presented in Model 3. Without a surprise, a positive and statistically significant relationship between QFII and CSR still exists. The evidence, again, supports our main finding. In unreported analysis, we also adopt alternative measures of CSR, such as the industry-adjusted CSR performance and the length of CSR reports, and in both cases, we find that QFIIs have a positive impact on CSR with the use of the firm fixed-effect model.

Weighted least squares approach
Our sample, an unbalanced panel data set, is unevenly distributed across years. It may be unreasonable to assume that every firm-year observation should be treated equally, and using a procedure that treats all of the data equally may provide a less precise estimation on data points since some may be given less influence than they should be, while others that have little influence are assigned with heavy weights. Bearing this in mind, we thus introduce the weighted least squares (WLS) methodology to maximise the efficiency of our parameter estimation. This approach assigns each data point a proper amount of influence over the parameter estimates. The result is displayed in Model 4 and it shows that the coefficient on QFII is positive and highly significant, suggesting that firms with QFIIs exhibit higher CSR scores when using a weighted regression model as an alternative robustness check.

Additional controls
We further test whether our main finding is sensitive to the inclusion of additional controls. First, recent empirical evidence shows that reduced agency problems and better corporate governance are both associated with superior CSR performance (Harjoto & Jo, 2011). In particular, Boubakri et al. (2016) find that firms cross-listed in multiple stock exchanges are subject to more rules on disclosure and corporate governance, and tend to have better CSR performance than those without cross-listing. To mitigate the concern that omitted variables may drive our results, we control for the effect of cross-listing on CSR performance by creating a dummy variable, Cross, which equals 1 if a firm is listed on multiple stock exchanges in a given year and afterwards, and 0 otherwise, and report the result in Model 5. The coefficient on QFII remains positive and statistically significant at the conventional level, supporting our main finding.
Second, we examine whether a firm's research and development (R&D) expenditures may affect the firm's investment in CSR activities. 15 The association between R&D intensity and subsequent CSR scores may go both ways. On the one hand, firms may choose to allocate fewer resources to CSR-related activities if there is a strong inclination towards R&D or innovative investments (Pavelin & Porter, 2008 advantages, leading to a positive link between R&D intensity and CSR (Hegde & Mishra, 2019;Padgett & Galan, 2010).
To account for the potential impact of R&D intensity on CSR, we generate a variable, indicated by RDIntensity and measured as the ratio of R&D expenditure to total assets in year t − 1, and repeat our analysis controlling for this variable.
The estimate is reported in Model 6 where it shows that RDIntensity has a positive and significant effect on CSR activities, implying that R&D-intensive firms tend to use CSR strategy as a means to recruit and retain talented people, such as research specialists and engineers, who can provide firms with critical intangible resources and capital market benefits, because social responsibility actions can fulfil stakeholder expectations and build a good reputation (Brammer & Pavelin, 2006). Note that the coefficient on QFII is still positive and statistically significant, suggesting that our main result is not affected by the inclusion of the R&D variable.
Third, we take firms' slack resources into account. According to slack resource theory, a firm's slack resource availability, such as excess cash balance, may have a positive effect on CSR performance (Waddock & Graves, 1997).
Although in our model specification, as shown in equation (1), variables such as ROA and FCF are included to take firm performance into account, excess cash accumulated by the firm may be even more important to socially based expenditure. Following Dittmar and Mahrt-Smith (2007) and Frésard and Salva (2010), we define excess cash, ExcessCash, as cash held by firms that is not committed for operations and investment activities scaled by total assets at year t − 1. In other words, excess cash is measured as the residual from the normal cash regression. 16 We repeat our analysis with inclusion of this variable into our model specification and display the result in Model 7. We find that the coefficient on excess cash is positive, but statistically insignificant (t = 1.54), indicating that slack resources available to the firm have no direct impact on CSR activities. Notably, however, the coefficient on QFII is still positive and statistically significant, reaffirming the positive influence of QFII on CSR ratings despite the inclusion of this additional variable.
Finally, as shown in our sample distribution foreign investors are clustered in the manufacturing sector, consistent with Liu, Bredin, Wang, and Yi (2014) who document that foreign funds are typically keen to invest in certain industries-for example, the manufacturing sector-as opposed to other sectors, such as real estate, construction, media and culture that generally need local knowledge. Thus, our result could simply reflect industry effects. Although the industrial dummy is included in all regressions shown in Table 4 to control for unobserved heterogeneity across industries, we further rule out such concerns by excluding firms in the manufacturing sector, which reduces our sample to 1,813 observations. Our result is reported in Model 8. As indicated clearly, the influence of QFIIs on CSR is still positive and statistically significant, implying that our main finding is not driven by a large number of QFIIs clustered in the manufacturing sector. It is worth noting that our finding further reveals that QFIIs play a more influential role in inducing appreciable changes in firms' CSR performance across broad industry sectors, thanks to the rapid expansion of the QFII Scheme in recent years. 17 In sum, our result is robust to a set of alternative research approaches and is insensitive to the inclusion of additional controls and sub-sample, further confirming that QFIIs play an important role in promoting CSR activities.

Endogeneity tests
As with all empirical studies in corporate finance, endogeneity is a possible concern. Firms with better environmental and social performance ratings may attract more foreign institutional investors due to their good reputation built through CSR activities. It might also be possible that the presence of foreign ownership and socially responsible conduct are simultaneously influenced by omitted variables. Further, the differences in observable firm characteristics between firms with and without QFIIs may result in potential concerns. Although the Granger causality test, as well 16 Please see Dittmar and Mahrt-Smith (2007) and Frésard and Salva (2010) for further details. 17 In particular, the initial quota that allowed QFIIs to invest in an individual stock was limited to a maximum of 20% of a firm's outstanding shares by CSRC. as firm-, industry-, and year fixed-effect regressions may, to some extent, mitigate these concerns, they cannot adequately address reverse causality problems and capture unobserved characteristics due to sample selection bias. In this section, we further address these concerns by adopting alternative research designs, including the propensity score-matching (PSM) method and the Heckman two-stage selection analysis.

Propensity score-matching approach
We first adopt the propensity score-matching method to examine the impact of foreign investors on CSR performance.
With replacement, we match the treatment group with the control group based on a variety of firm characteristics (Size, Age, Lev, and ROA) as well as the year and industry dummies, and employ the PSM technique with calliper set at 0.01. We identify the average treatment effect on the treated (ATT), in which the 'Treatment' represents QFII. ATT is the average difference between the CSR ratings of firms with QFIIs and their counterfactual peers' CSR ratings. After matching, a balancing test is carried out to examine whether the mean of each covariate significantly differs from the treatment to control groups.
A propensity score analysis for treatment firms and non-treatment firms is presented in Panel A of Table 6. In Section A, it shows that the CSR performance of firms with QFIIs is approximately 1.40 points higher than that of firms without QFIIs. Section B indicates that the sample is well balanced, as the firm characteristics of the control group are not statistically different from those of the treatment group in all panels. 18

Heckman two-stage estimator
Our sample is constructed based on whether firms disclose CSR or sustainability reports and receive CSR performance scores, which may potentially introduce sample selection bias because not all firms disclose their CSR information and sustainability reports (Kong, Yang, Liu, & Yang, 2020). To alleviate such concerns, we employ the Heckman (1979) twostep procedure to check whether our result is robust to this selection bias. To perform this test, we first carry out a probit regression model, as shown in equation (2), to estimate the likelihood that firms issue CSR reports and receive CSR performance scores, to obtain the Inverse Mills Ratio (known as Lambda), the self-selection parameter, and in the second-stage regression, we insert Lambda in the model specification as shown in equation (3): where CSRD is an indicator variable set to 1 if a firm issues a CSR report and receives a CSR rating in a given year, and 0 otherwise. Mandatory is used to measure the probability of whether a sample can be observed, and is defined as a dummy variable equal to 1 if a firm is mandated to disclose a CSR report in a given year, and 0 otherwise (Chen, Hung, & Wang, 2018;Kong et al., 2020). 19 According to prior research, Mandatory affects firm-level CSR but has no direct impact on the presence of QFIIs. Control is the same set of control variables included in equation (1) that are known to have a potential impact on CSR. 18 We also identify the ATT when the 'Treatment' is QFII_high and QFII_distant, and our finding remains unchanged. 19 Note that firms listed on the SSE Corporate Governance Sector and cross-listed are required to disclose their CSR reports. In addition, firms included in the Shenzhen 100 Index are required to issue their sustainability reports. Notes: Panel A reports the results of a propensity score matching (PSM) routine for treatment firms and non-treatment firms from 2009 to 2017. We match firms using a propensity score matching algorithm with calliper set at 0.01 and an array of firmspecific characteristics (firm size, firm age, leverage, and ROA). The average treatment effect on the treated (ATT), in which the 'Treatment' stands for QFII, is reported in Section A. Section B displays the univariate balanced test for pairs of treatment and control firms after matching. All matching criteria are all in year t − 1 with year and industry dummies included. Panel B presents the results based on Heckman two-step regressions. In the first stage, a probit model is adopted to predict the presence of CSR performance, where a dummy variable, CSRD, set to 1 if a firm issues a CSR report and receives CSR rating in a given year, and 0 otherwise, is regressed on Mandatory and Control to estimate the Inverse Mills ratio, namely Lambda. Mandatory is a dummy variable equal to 1 if a firm is mandated to disclose a CSR report in a given year and 0 otherwise. In the second stage, Lambda is included as an additional independent variable in equation (1). Due to space limitations, this panel does not display the control variables. In both regressions, independent variables are lagged by one year, with controlling for industry and year fixed effects. Detailed variable definitions can be found in Appendix B. Standard errors are corrected for heteroscedasticity and clustered at the firm and year level. T-statistics (or Z-statistics) are reported in parentheses. *** , ** , and * indicate statistical significance at the 1%, 5%, and 10% level, respectively.

TA B L E 6 Endogeneity tests
Panel B of Table 6 reports the regression results of the Heckman two-stage analysis. In the first-stage regression, the coefficient on Mandatory is positive and highly significant, indicating that firms subject to the mandatory CSR reporting policy are more likely to issue their sustainability reports and receive CSR performance ratings. In the second-stage regression with Lambda included, our results show that the coefficient on QFII is positive and statistically significant (t = 2.92), consistent with our main finding. The magnitude of the coefficient on QFII (1.43) is slightly lower than that in Model 1 of Table 4 (1.67), implying that the selection bias slightly amplifies the impact of QFIIs on CSR performance. More importantly, the positive association between the presence of foreign institutional investors and CSR performance is unaffected using the Heckman two-stage analysis. 20

WHY DO QFIIS PUSH FIRMS TOWARDS BETTER CSR PRACTICES?
In this section, we aim to answer the question of why QFIIs push firms to adopt high CSR practices by looking into two avenues: social awareness and physical distances of QFIIs. The former reflects foreign investors' social awareness, and the latter may imply an additional degree of information asymmetry that QFIIs may bear.

The social awareness motive: Regulatory quality of countries
When assessing the positive effect of foreign investors on CSR performance, it is necessary to review QFIIs' legal and regulatory quality backgrounds that may reflect their distinctive social preferences. Liang and Renneboog (2017) assert that countries under different legal regimes handle concerns and conflicts between firms, communities and other stakeholders differently; regulatory quality and institutional backgrounds can, to a large extent, explain the variation in the corporate social practices across countries. Dyck et al. (2019) maintain that institutional ownership, especially those from countries with a strong community belief in the importance of environmental and social (E&S) issues, is a driving force behind firms' E&S performance. In addition, according to Worldwide Governance Indicators (WGI), institutions in countries with strict law and regulations can effectively enhance the stringency of environmental regulations, labour regulations, business regulatory environment, and address various stakeholder concerns in terms of unfair competitive practices, access to product markets, labour rights, minimum wages, hiring and firing practices, and price control (Mavragani, Nikolaou, & Tsagarakis, 2016). Furthermore, well-governed countries largely depend on regulation-based mechanisms that place restrictions on ex-ante behaviours (La Porta, Lopez-de-Silanes, & Shleifer, 2008;. As a consequence, foreign investors from countries with high regulatory quality are more likely to exhibit their stronger social preferences, and when they invest in foreign markets they may carry out their high social behaviour in incorporating environmental and community engagement into firms that they invest in, thus pushing towards higher CSR behaviours of the firms.
To examine whether foreign investors transplant their social awareness to the firms they invest in, we divide QFIIs into two groups according to the regulatory quality of countries where they are domiciled. Specifically, we obtain the country's regulatory quality scores from the World Bank and calculate the median value for each year. 21 We then generate two dummy variables. QFII_high is assigned with a value of 1 for firms having a foreign investor from a country with a regulatory quality score equal to or above median value in a given year, and 0 for firms either having no foreign investors or having foreign parties from a country with a regulatory quality score lower than median value.
Similarly, QFII_low receives a value of 1 for firms having foreign institutional investors from a country with regulatory 20 We have also adopted the Heckman two-stage analysis to address the selection bias concerns with respect to the link between social awareness and geographical distance of QFIIs and CSR performance by replacing QFII with QFII_high and QFII_low, and QFII with QFII_distant and QFII_close in equations (2) and (3), respectively. Our main finding continues to hold. quality score lower than median value, and 0 for firms either having no foreign institutional ownership or having foreign investors from a regime with regulatory quality score equal to or above median value. In the case of multiple foreign investors within a firm in a given year, we identify the dominant institution according to the total percentage of shares held by QFIIs. 22 By replacing QFII with QFII_high and QFII_low in equation (1), we obtain the estimates reported in Model 1 of Table 7. Notably, the coefficient on QFII_high is positive and statistically significant, suggesting that the positive effect of foreign institutional investors on CSR performance is mostly attributable to QFIIs domiciled in highly regulated countries where the social awareness and CSR standards are higher than those from countries with relatively low regulatory quality. Therefore, our results reveal that the legal and regulatory quality background of QFIIs can be viewed as an essential channel that transplants social standards to the firms that they invest in, and offshore owners with high social awareness background act as a key driver of the initiative of CSR practices.

The risk resolution motive: Physical distance of QFIIs
Alternatively, foreign investors' push for CSR engagement may stem from the idea of investment uncertainty reduction that CSR engagement may bring. Investing in a foreign country is associated with additional uncertainty and higher levels of risk due to increased information asymmetry (Barmeyer & Mayrhofer, 2008;Gehrig, 1993;Oh et al., 2011).
Specifically, Kang and Kim (2008) find that information asymmetry arising from geographic proximity is a key determinant of domestic large acquirers' corporate governance practices in their target firms, implying that investors from geographically distant markets suffer from more severe information asymmetries relative to investors domiciled in physically close countries. Furthermore, as indicated in Leuz, Lins, and Warnock (2008) and Cai, Lee, Xu, and Zeng (2019), foreign institutional investors are often assumed to exhibit information disadvantages relative to domestic investors who have greater local knowledge and access to a wider variety of additional information from local sources.
As such, it is plausible to assume that QFIIs pressure managers to promote CSR activities as a way of reducing information asymmetry that they bear due to geographical distances, especially those from remote countries, because the empirical evidence shows that effective use of CSR strategies is negatively associated with information asymmetry (Cho, Lee, & Pfeiffer, 2013;Cui, Jo, & Na, 2018).
To assess the uncertainty resolution incentive, we follow Kang and Kim (2010) and measure the degree of information asymmetry using the natural logarithm of geographical distance between the capital cities of QFIIs' countries of domicile and the capital city of China as a proxy. We then divide QFIIs into those from remote-distance countries (equal or above-median physical distance) and near-distance countries (below-median physical distance). Accordingly, two dummy variables are created with QFII_distant being a binary variable equal to 1 for firms with foreign institutional investors from remote-distance countries in a given year, and 0 for firms either without foreign institutional ownership or with foreign investors from near-distance countries in the same year, and similarly QFII_close being a dummy variable set to 1 for firms that receive foreign investments from near-distance countries in a given year, and 0 for firms either without foreign institutional ownership or with foreign investors domiciled in remote-distance countries in the same year. The estimate of equation (1)  The regulatory quality scores of Hong Kong (2.06) and Canada (1.85) are higher than the median level (1.67), thus a high regulatory system, while the regulatory quality score of Kuwait (−0.15) is lower than the median value, thereby a relatively inferior governance system. Therefore, the dominant QFII in Jiangling is identified as the high-regulatory-quality system because 2.78% is greater than 1.03%.

TA B L E 7
The effects of regulatory quality and geographical distance of QFIIs' countries on CSR and the mechanism that QFIIs use to push for CSR changes  (1) and (2), and ownership characteristics of QFIIs in Models (3) and (4) on the CSR performance of firms invested. See Appendix B for detailed variable definitions. In all models, independent variables are lagged by one year. Standard errors are clustered at the firm and year level and are corrected for heteroscedasticity. T-statistics are reported in parentheses. *** , ** , and * indicate statistical significance at the 1%, 5%, and 10% level, respectively.
additional risk arising from information asymmetry that they bear due to the geographical location from the target firms.
Together, our findings support that both social awareness and risk resolution are the driving forces for offshore owners to urge firms that they invest in to propel environmental and social policies.

WHAT MECHANISMS DO QFIIS USE TO PUSH FOR HIGH CSR?
The literature suggests that large and key shareholders can exert influence on corporate policies through proposing and voting, or by appointing directors on the board or shareholder activism (Lee & Lounsbury, 2011;Oh et al., 2011).
Similarly, if foreign institutional investors own a significant amount of a firm's shares, they are entitled to vote in the shareholder meeting (Huang & Zhu, 2015

EXTENDED ANALYSIS
As discussed earlier, the presence of foreign investors plays a part in facilitating socially responsible policies and leads to an improvement in CSR ratings. In this section, we extend this line of analysis by assessing whether the increase in CSR ratings associated with the presence of QFIIs has real consequences on firms' business strategies. Prior studies document that superior CSR performance may have implications on firm performance (Cook, Romi, Sánchez, & Sánchez, 2019;Gregory, Whittaker, & Yan, 2016;Lins, Servaes, & Tamayo, 2017) and access to finance (Cheng, Ioannou, & Serafeim, 2014). Next, we extend the study by exploring the incremental effects of CSR driven by the presence of foreign institutional investors from these two aspects.
First, we look into the interplay between QFIIs and CSR performance on firm profitability. Prior empirical studies reveal that socially responsible activities and practices are consistent with value maximisation (Ferrell et al., 2016;Lins et al., 2017). Indeed, CSR engagement is found to exert a significant effect on corporate strategies motivated by the economic and social benefits of these philanthropic activities. Grounded on stakeholder theory, the success of an organisation relies heavily on its ability to develop a mutually respectful and trustful relationship with different stakeholder groups (Deng et al., 2013). Inspired by these studies, we empirically investigate whether firms with QFIIs can obtain profit-related benefits from the enhanced CSR performance, by specifically focusing on ROA, a measure of a firm's operating performance as a critical indicator for monitoring managers' earnings capabilities, with the following specification: where ROA stands for operating performance, measured as earnings before interest and taxes scaled by the book value of total assets (Ding, Ferreira, & Wongchoti, 2016). The incremental effect of CSR performance through the presence of foreign stakeholders on firm profitability is captured by the interaction term between QFII and CSR, which is our main interest. We expect this effect to be positive according to the prior literature that the enhanced CSR performance should lead to better profitability. Following Boubakri et al. (2016) and Wang and Li (2016) Second, we examine the interaction between QFIIs and CSR performance on access to finance. From the perspective of capital constraints, ethical stakeholder engagements lead to better access to finance as a result of more transparency in disclosures regarding operations and financial position (Cheng et al., 2014). We thus expect that an increase in CSR driven by the presence of foreign institutional investors may have an incremental effect on firms' access to finance. Following Kaplan and Zingales (1997) and Cheng et al. (2014), we use their regression coefficients to construct a KZ index for each firm-year observation as a dependent variable. A higher value of KZ_index suggests that the firm has more capital constraint. We then re-estimate equation (4) Kaplan and Zingales (1997) for details on the construction of KZ_index.

CONCLUSIONS
In this study, we investigate the role of foreign institutional investors on firms' CSR performance and reporting. Using a panel of publicly traded firms in China between 2009 and 2017, we provide evidence that QFIIs push firms towards high CSR engagement. In addition, we find that firms with foreign investors are more likely to comply with the GRI guidelines, and their reports on sustainability tend to be longer. Furthermore, our results reveal that such a positive effect is more pronounced in firms with low initial CSR performance than those with high initial CSR performance at the time when QFIIs enter the sample, suggesting that the scope of the impact from QFIIs varies, depending on the initial CSR level of a firm. Our finding is robust to several additional approaches, including the Granger causality test, a firm fixed-effect model, and weighted least squares, and is insensitive to controlling for effects of cross-listing, R&D intensity, and slack resources, as well as a sub-sample excluding firms in the manufacturing sector. Importantly, our finding holds after endogeneity concerns are carefully addressed, based on alternative research designs, such as propensity score matching and the Heckman two-stage selection analysis.
We further look into the motives of QFIIs and reveal that both social awareness and geographical distance are the driving forces behind this positive impact. In particular, we find that this positive effect is more pronounced in firms with foreign investors domiciled in countries with high regulatory quality or from geographically remote distance. In addition, our evidence suggests that the mechanism underlying such positive effect is through the ownership of QFIIs, especially when the QFII is among the top ten of the largest shareholders, which provide foreign investors with sufficient voting power to push firms towards stronger CSR performance. Finally, in the extended analysis, we find that the improvement in CSR associated with the presence of QFIIs results in better operating performance and easier access to finance.
Overall, this study offers valuable implications for policymakers in emerging markets where corporate social conduct practices are relatively low. Specifically, our finding suggests that foreign institutional investors, especially those from nations with high regulatory quality, or those from geographically distant countries, may effectively help integrate financial or human resources and exert a positive influence on social strategies of firms in which they invest.
Therefore, other emerging markets should consider introducing a similar QFII scheme in helping their domestic firms to achieve higher CSR practices in accordance with international standards. -Strategy M2 Impact of climate change, social issues, and change in macroeconomic environment on corporate sustainability.

M3
Impact of products or business services on society and the environment.

M5
Long-term and short-term CSR plans.
-Governance M6 Disclosure of firm, industry, and social environment information.

M7
CSR values and codes of conduct.

M8
The monitoring of CSR implementation.

M9
The management process in terms of corporate social, economic, and environmental issues.

M10
Institutional settings regarding information disclosure.

M11
Assessment of sustainability-related risks.

M12
Disclosure of anti-corruption and anti-commercial bribery information and related administrative controls.

M13
CSR promotion among affiliated enterprises and subsidiaries.

M15
Stakeholder communication; feedback from stakeholders.

M16
Review and comments from stakeholders.

C1
Disclosure of annual total revenues, profits, and dividend distribution plans.
-Economic performance C2 Disclosure of percentage change in total revenues, profits, and dividend distribution plans on a year-on-year basis.

C3
Sales volume of services and products; market share of services and products; innovation.
-Labour and human rights C4 Gender structure; age distribution; total number of employees; total number of temporary employees; formal employment contracts.

C5
Employee training and development.

C6
Employee health and safety; protective equipment.

C7
Employee complaints and relevant solutions; declaration on the prohibition of the use of child labour; equal pay for equal work.

C8
Employee compensation; holiday entitlement; other benefits in addition to remuneration.

C9
Labour union; staff entertainment; family care; employee satisfaction surveys.

C10
Employee engagement in sustainability.
-Environment C11 Annual environmental protection investments; environmental management system.

C12
Measure, record, and report the source of contamination; pollution control.

C14
Identify the source of greenhouse gas emission; climate change; greenhouse gas emission reduction; reduce or avoid the adverse effects of climate change.

C16
Integrate social and environmental concerns in business operations.
-Customers C17 Quality management certification; technological innovation in services and products.

C18
Customer relationship management; customer satisfaction survey.

C19
The percentage of the product passing all quality requirements without rework; product safety; recycling mechanism.

C20
Customer complaints and related solutions.

C21
Customer privacy protection.

C22
Product instructions; consumer rights.
-Community participation and development C23 Total social donations.

C25
Rules and regulations.

C26
Job creation and the number of employees recruited by year.

C27
Participation in technological projects; cooperation with universities or research institutes.

T1
The comprehensive coverage of stakeholders' responsibilities.
-Information coverage T2 Negative information disclosure and challenges.

T4
Coverage of CSR ratings; CSR reporting; CSR ranking.

T6
Effectiveness of innovation.
-Reliability and transparency T7 Disclosure level of stakeholders' suggestions.

T9
The certificate authority of the third-party agencies.

T10
Readers' suggestions and feedback.
-Normalisation T11 The time frame of the report; coverage; information regarding participants and producers.
-Standard T12 Standards and control groups.
-Information effectiveness T14 Multiple language versions of CSR reports.

T15
Access to CSR reports.