Environmental, Social, and Governance Performance, Platform Governance, and Value Creation of Platform Enterprises

Abstract

Under the concepts of sustainable development and a sharing economy, the ESG performance of platform enterprises has played a significant role in measuring the operating status and responsible investment of platform enterprises. Platform enterprises have different typical characteristics from traditional enterprises. The mechanisms of ESG and financial performance needs to be further explored. The empirical analysis finds that: (1) the ESG performance of platform enterprises and its S index and G index has a positive impact on corporate financial performance. (2) Media attention plays a positive moderating role between the ESG and ROA. (3) Platform data governance and platform reputation governance are two internal and external paths for platform enterprises’ ESG performance to improve financial performance. (4)There is heterogeneity in the relationship between ESG and ROA in terms of platform enterprise scale and platform type. Based on the above conclusions, this paper provides reference experience for the ESG governance and value creation of platform enterprises.

1. Introduction

ESG is a three-dimensional coordinate system, including three dimensions, namely E, S, and G [1,2,3]. The concept of ESG has received much attention since it was put forward. For example, the Shanghai Stock Exchange requires the disclosure of ESG information, and many enterprises have begun to actively disclose ESG reports or annual reports containing ESG information. Scholars have already made relevant studies on the role of ESG in traditional enterprises. Sun Dong and others believe that ESG rating results are important factors to consider when making decisions [4]. Qu Xinjiang (2024) also proposes that investors should consider the ESG performance of an enterprise in addition to its financial factors before judging whether an enterprise can be invested in [5]. With the arrival of the information age and the rapid development of the platform economy, Internet platform companies face greater public opinion monitoring and information disclosure requirements, and companies are increasingly emphasizing non-financial performance governance modes, including regularly released ESG reports [6,7,8,9]. In the context of today’s economy shifting from rough growth to the pursuit of sustainable green development, platform enterprises are required to actively assume ESG responsibilities, and whether they can realize the improvement of corporate performance is their most critical motivation [10,11,12].

Unlike traditional enterprises, platform enterprises are an emerging form of enterprise. In terms of enterprise characteristics, platform enterprises are characterized by their large volume, rapid development, and strong cross-network externalities, which enable them to have a wider audience and social impact. In terms of industry characteristics, platform enterprises span a wide range of industries and have become an important part of industrial upgrading and people’s daily lives. In terms of the organizational form of enterprises, platform enterprises a have more complex organizational ecology due to the multiple participants involved. The lack of corporate environmental, social, and governance responsibilities faced by platform enterprises is characterized by diversity, and the rights and interests of consumers need to be safeguarded. For example, in reality, the illegal collection of personal information, big data-enabled price discrimination against existing customers, platform mandatory binary choice, rental platform operating explosion, community group competing battle, content vulgarity, and other industry problems occur frequently. As recently as the beginning of 2022, the State Administration for Market Supervision and Administration announced 13 penalty decisions in cases of illegal implementation of operator concentration in the Internet field without declaring these in accordance with the law, and administrative penalty decisions were made on large Internet platforms, such as Ali, Tencent, and Beili Beili. Therefore, it is imperative for platform companies to actively engage in ESG to create a favorable online environment and a culture of responsibility. In order to safeguard the rights and interests of consumers, the intrinsic drivers of platform enterprises’ social responsibility undertaking and ESG responsibility enhancement deserve to be further explored.

Similar to the classical studies on traditional CSR, what kind of relationship exists between both ESG and the financial performance of platform firms is the key question for the aforementioned intrinsic drivers. At present, only a few studies focus on the relationship between the two, most of which show that there is a positive correlation between them [13,14,15,16,17,18,19]. Most of the existing research is based on the perspective of platform corporate social responsibility, and most of its indicators are constructed according to the dimension of platform corporate social responsibility. Compared to platform corporate social responsibility, ESG has a more convergent normative quantitative system according to the indicators constructed in E, S, and G, but there is little research on it. On the one hand, from the perspective of platform enterprise ESG performance, platform firms’ improved ESG performance is conducive to safeguarding the interests of various stakeholders, such as consumers, firms, and society, and to guiding the correct public habits and values, which may enhance financial performance through signaling and reputational effects. On the other hand, unlike traditional enterprises, the platform business model and business attributes are characterized by rapidity, networking, a wide scope, and high impact. Are the cost inducements in traditional CSR and financial performance studies still applicable to the highly intensive production and operation mode of platform enterprises? Therefore, whether the effect of ESG performance of platform enterprises affecting financial performance and the process of value creation is the same as that of traditional enterprises deserves further exploration. With the continuous development of the platform economy, the environmental, social, and governance deficiencies of platform enterprises have attracted widespread attention. In the past, scholars have achieved rich research results on the relationship between the two, but in view of the typical characteristics of platform enterprises different from traditional enterprises, the relationship between the two under the background of a sharing economy still needs further in-depth discussion.

2. Theoretical Analysis and Research Hypotheses

2.1. ESG and the Financial Performance of Platform Companies

According to the two-sided market theory, platform enterprises connect the supply side and the demand side. In the face of a more complex organizational environment and multiple participants, good ESG performance can help enterprises maintain competitive advantages and expand the market at all times. Based on stakeholder theory, all stakeholders of platform enterprises are in an Internet ecosystem. In order to realize the platform enterprises’ own interests and the harmonious development of the ecosystem, platform enterprises have to consider the interests of various stakeholders, who are usually more inclined to choose enterprises with higher ESG level for cooperation [20,21]. The implementation of ESG is conducive to improving platform reputation governance for existing businesses, building product advantages, and ultimately improving the financial performance of enterprises [22,23,24,25,26]. Therefore, good ESG performance can not only improve the existing opportunities of enterprises but can also improve the potential opportunities of enterprises [27]. In addition, in order to alleviate this asymmetry, it is necessary to carry out good signal transmission. Every company has the opportunity to show or not show its true quality to the outside world, and high-quality companies are more motivated to send signals that help the outside world distinguish between companies. The extensive public characteristics of the platform enterprises are more convenient to alleviate the information asymmetry through signal transmission. Platform enterprises can eliminate the information asymmetry between the company and consumers, investors, public groups, etc., and affect financial performance by transmitting good information on fulfilling social responsibilities to various stakeholders [28,29].

From the perspective of each dimension of ESG, enterprises in the E dimension often engage in energy conservation, environmental protection, and a circular economy through R&D innovation, which not only conforms to the concept of sustainable development, but also can open up new potential markets and occupy a larger market share, so as to improve the financial performance of enterprises [30,31]. Under the S dimension, it is easier for enterprises to attract high-quality employees by improving the welfare of employees in enterprises, and the efficiency of enterprises will also be improved [32]. For platform enterprises, paying attention to the welfare of more employees, the employees of the enterprise will have more stickiness to the enterprise, and it is easier for the enterprise to maintain a positive atmosphere, thus improving the efficiency of the enterprise. Under the G dimension, enterprises can improve their work efficiency by improving their self-management, so as to improve their financial performance [33]. However, based on the cost–benefit theory, it can be seen that enterprises first need to invest a lot in terms of cost in order to obtain later benefits. Therefore, the investment of enterprises in ESG performance may affect the financial performance of enterprises [34].

In traditional enterprises, the performance of enterprise ESG has a significant positive effect on promoting the financial performance of listed companies [35,36,37]. In platform enterprises, stakeholder responsibility is positively related to the current financial performance of the enterprise [17]. The fulfillment of social responsibility by platform enterprises significantly improves financial performance and enterprise value [18,19]. Compared with traditional enterprises, platform enterprises have more active users, more attention, richer network connections, faster market response, and wider social influence. In addition to the economies of scale and economies of scope similar to traditional enterprises, platform enterprises are more obvious in the intensification of production and operation. Compared with traditional enterprises, the role of cost incentives in the relationship between the two has weakened. Therefore, the active implementation of ESG by platform enterprises can improve their financial performance through the signal transmission effect, cost reduction effect, and comprehensive effect of various dimensions.

H1. 

The ESG performance of platform enterprises has a positive impact on their financial performance.

2.2. Moderating Effects of Media Attention

From the perspective of the supervision effect, the media has a supervision effect on the ESG of platform enterprises [38,39]. Luo Jinhui found that high-level media coverage as a monitoring mechanism can effectively reduce the company’s dual agency costs (Luo Jinhui, 2012) [40]. For platform enterprises with good performance of ESG, it is easier to obtain the media’s favor and receive more reporting. More reports will gain more stakeholder attention, increasing the opportunity cost of the enterprise [41,42].

From the perspective of expectation theory, individual decision-making behavior is based on cognition, and the judgment of satisfaction comes from the consistency of expectations. When the ESG performance of platform enterprises is consistent with the expectations of stakeholders for platform enterprises, the satisfaction of stakeholders is higher. The media can improve the attention of stakeholders to the ESG of platform enterprises through effective ESG information transmission, and then improve the satisfaction of stakeholders, so as to strengthen the role of the ESG performance of platform enterprises in financial performance [43,44].

From the perspective of information asymmetry, media coverage of enterprises is conducive to reducing the information gap between enterprises and stakeholders, and in alleviating the possible lag effect due to information transmission in the impact of enterprise ESG on financial performance [45,46,47]. With the rapid development of information, platform enterprises have a huge user group and are more concerned by the media. Good ESG performance will alleviate the public’s poor information about the company and enhance the user’s sense of trust [48]. In addition, the media’s coverage of enterprises can have a strong ability to guide public opinion, which enlarges the effect of platform enterprise ESG and plays an important role in stakeholder satisfaction [49].

H2. 

Media attention plays a positive moderating role in the impact of platform enterprise ESG on enterprise performance.

2.3. Mediating Effects of Platform Reputation Governance and Platform Data Governance

Reputation theory refers to the image and evaluation of enterprises in society, which can be used as an incentive mechanism to have an incentive effect on enterprises. A good corporate reputation will be recognized by stakeholders, thereby improving corporate performance. The improvement of corporate performance will encourage enterprises to actively enhance their reputation and gain the favor of more stakeholders. Corporate reputation is the unique competitiveness and resources of an enterprise, and a good corporate reputation can tap the unique competitiveness of enterprises and increase the potential opportunities of enterprises to create value [50,51].

Corporate reputation often needs to be transmitted by certain media. The development of Internet technology makes media reports play a more important role. From the perspective of corporate reputation theory, media coverage can use its advantages to truly reflect the reality of the company, thus promoting the reputation of the company [52]. Positive reports in media reports are conducive to affecting the image of enterprises and improving the reputation of enterprises. A good reputation means that potential investors are more likely to respond positively to an enterprise. In addition, platform reputation governance is conducive to reducing the information asymmetry of enterprises and conveys external signals that management and stakeholders care more [53,54,55]. The media enhances the goodwill of stakeholders to the enterprise by sending a good signal to the outside world, thereby increasing the positive correlation between ESG performance and financial performance.

H3. 

Platform reputation governance plays an external intermediary role in the relationship between platform enterprise ESG performance and enterprise financial performance.

Platform data governance has become an important part of digital technology and industrial platforms, such as the Internet and mobile devices. For example, social media platforms can promote social harmony through the internal management of false information and hate speech, and e-commerce platforms can promote fair competition by protecting consumer rights and preventing the sale of counterfeit goods. However, with the increasing centralization of the platform economy, especially the emergence of super-large platforms, people are increasingly concerned about the negative effects of platforms, such as the abuse of market dominance, the infringement of user data privacy, and improper use of artificial intelligence algorithms.

Most platform stakeholders are often unable to restrain the lack of corporate social responsibility. Platform enterprises lacking internal management and supervision may lead to negative effects and even cover up the good signal of responsible ESG. On the one hand, platform data governance based on rules and mechanisms guides enterprises to fulfill ESG and accept social supervision. On the other hand, it regulates corporate behavior and reduces negative signals. Platform data governance requires enterprises to carry out a technical evaluation, security audit, risk prevention and control, etc., which helps to reduce the degree of stakeholders being limited by information asymmetry and to enhance the confidence of stakeholders in the enterprise. This belongs to an internal intermediary. In addition, platform data governance is based on the government’s requirements for the internal responsibilities of platform enterprises, and the performance of the ESG of platform enterprises comes from the comprehensive evaluation of corporate environment, society, and governance responsibilities [56,57,58]. Therefore, there is an inherent logical consistency between the ESG performance of platform enterprises and platform data governance. The data governance level of the platform represents the governance ability of the platform, and the high governance ability is a positive signal for the platform enterprises. Platform enterprises actively fulfill ESG and send positive signals through the level of platform data governance, which not only meets the government’s requirements for platform enterprise responsibility, but also makes the stakeholders of platform enterprises more confident in platform enterprises, thus affecting the financial performance of such enterprises [59,60].

H4. 

Platform data governance plays an internal intermediary role in the relationship between ESG performance and financial performance of platform enterprises.

3. Technology

3.1. Data and Sample

This study mainly combines the connotation of platform enterprises, refers to the platform business classification in Li Ying, Song Yan et al. [17,18] and the ‘Internet and related service operation in 2021’ and ‘Internet platform classification and grading guide (draft)’ issued by the Ministry of Industry and Information Technology, selects A-share listed companies in 2013–2022 as the samples, and processes the data as follows: (1) Keywords, such as ‘platform’, ‘Internet’, ‘e-commerce’, ‘social entertainment’, and ‘intermediary’ were searched for. Through the business scope and main business of the enterprise, the platform enterprise was preliminarily screened; (2) we eliminated ST and * ST companies; (3) we excluded 2013 IPO companies; (4) we excluded companies with missing data and outliers; (5) we excluded financial industry enterprises. By comparing the selected platform enterprises with their annual report information, 1060 sample observations were finally obtained. This paper performs winsorize tail reduction on continuous variables at 1% and 99% levels. The data in this paper passed the VIF test.

3.2. Measurements

Explained variable. Financial performance. There are generally two types of financial performance evaluation. One is accounting-related indicators, such as ROA [61], total asset net interest rate, and sales net interest rate. The second category is indicators related to the capital market, such as Tobin’s Q. This paper selects the ROA data in the CSMAR database as a financial performance indicator.

Explanatory variable. Enterprise ESG performance. ESG performance is generally measured by three methods: the first draws on the results of relevant rating agencies, such as Huazheng; the second establishes an index system around the three indicators of ESG. The third involves issuing questionnaires and analyzing them. This paper chooses Huazheng data to measure ESG performance, mainly because the database contains a large number of sample enterprises and the starting year is earlier. Huazheng ESG performance includes 3 first-level pillar indicators, namely E, S, and G, as well as 16 s-level thematic indicators, 44 third-level issue indicators, and nearly 80 fourth-level bottom indicators.

Moderating variables. Media attention, drawing on the research of scholars, such as Wen and Zhou [62], Yuan and Xiong [63], and using the total number of online media and newspaper financial news of China Research Data Service Platform to measure media attention. The data of the China Research Data Service Platform follow the hot spots and usually have characteristic data.

Intermediary variable. Platform reputation governance. Drawing on the research of scholars, such as Shen Hongtao and Li Zhibinand Feng Jie [64,65,66], we use the Janis–Fadner coefficient (J $-$ F) to construct platform reputation governance indicators.

The following are the J $-$ F coefficients:

$$\mathrm{J}-\mathrm{F} \mathrm{c}\mathrm{o}\mathrm{e}\mathrm{f}\mathrm{f}\mathrm{i}\mathrm{c}\mathrm{i}\mathrm{e}\mathrm{n}\mathrm{t}\left\{\begin{array}{cc}{\displaystyle \frac{{\mathrm{e}}^2-\mathrm{e}\mathrm{c}}{{\mathrm{t}}^2}}& \left(\mathrm{i}\mathrm{f} \mathrm{e}>\mathrm{c}\right)\\ {\displaystyle \frac{\mathrm{e}\mathrm{c}-{\mathrm{c}}^2}{{\mathrm{t}}^2}}& \left(\mathrm{i}\mathrm{f} \mathrm{e}<\mathrm{c}\right)\\ 0& \left(\mathrm{i}\mathrm{f} \mathrm{e}=\mathrm{c}\right)\end{array}\right.$$

where e and c are the number of positive and negative media reports, respectively, and t is the sum of the two. The value of the coefficient is between −1 and 1. The larger the value, the more positive the report on the enterprise in the year.

In terms of platform data governance, the selected indicators come from the keywords in the “Guidelines for the Implementation of the Main Responsibility of Internet Platforms (Draft for Comments)” drafted by the General Administration of Market Supervision (GAMS). Through the jieba text analysis in Python, the keyword frequency of platform data governance in the annual report of listed companies is used as the proxy variable of platform data governance, as shown in

Table A1. Platform data governance.

Platform Data Governance Keywords
1. Fair cCompetition demonstration	Fair competition; non-public data; affiliated platforms
2. Equal governance	Equal governance; equity; non-discrimination
3. Open ecology	Open Ecology; facilitation
4. Data management	Data management; data security
5. Internal governance	Internal governance; platform compliance; platform legal status
6. Risk assessment	Risk assessment; violation of consumer rights; content audit; ad-targeted recommendations; safe and stable operation of platforms
7. Risk prevention and control	Risk prevention and control; risk assessment reports; emergency safeguards
8. Security audit	Security audit
9. Promoting innovation	Blaze new trails
10. Information verification, recording and disclosure	Information verification; information record; information disclosure; information integrity; information confidentiality; information availability
11. User management within the platform	User management; termination of service; violation of law; restriction of service provision; assistance in investigation; assistance in pursuing responsibility
12. Platform content management	Content management; warning; restriction of publication; Cessation of transmission; elimination of information; suspension of updates; closure of accounts; reporting complaints; infringement of legal rights and interests; seeking illegal gains
13. Prohibition and restriction of sale control	Prohibition of sales; restriction of sales; platform audit; daily inspection
14. Service agreements and transaction rules	Service agreements; transaction rules; rule fairness; rule justice Disclosure of rules
15. Credit evaluation	Credit rating; Internet integrity
16. Antitrust	Anti-trust legislation
17. Anti-unfair competition	Anti-unfair competition; false information; misleading information
18. Data acquisition	Data acquisition
19. Rule of law	Laws and regulations; fundamental rights of citizens; legal rights and interests of enterprises; algorithmic regulation; social supervision
20. Regulation of price behavior	Code of pricing conduct; compliance with pricing laws; compliance with pricing regulations; price discrimination; price gouging; dumping at low prices; luring consumers; compliance risk
21. Advertising code of conduct	Advertising code of conduct; illegal advertising; forewarning; post-processing
22. Intellectual property protection	Intellectual property protection
23. Prohibition of pyramid schemes	Prohibition of pyramid schemes; pyramid schemes; pyramid schemes information
24. Governance of black and gray products on the Internet	Healthy Internet environment; orderly Internet environment; information sharing; consultation and notification; specialized remediation; case assistance; Internet black and gray production governance
25. Network security	Cybersecurity; emergency preparedness; cybersecurity protection
26. Data security	Data security; data security protection
27. Privacy of natural persons and protection of personal information	Privacy; personal information protection; information leakage; information tampering; information loss
28. Consumer protection	Consumer protection; complaints; reporting; online dispute resolution; deceiving consumers; misleading consumers; internal oversight inspections; safeguarding personal safety; safeguarding property; safety and security
29. Protection of operators on the platform	Intra-platform operator protection; reasonable restrictions; reasonable conditions; reasonable costs; complaints; access to remedies
30. Protection of workers	Worker protection; physical and mental health; safe working environment; fair remuneration; reasonable remuneration; personal accident protection; employment
31. Protection of special groups	Protection of special groups; protection of minors; personality development; legal rights and interests; protection of the elderly; protection of persons with disabilities; technical support
32. Environmental protection	Environmental protection; practicing saving; opposing waste; efficient utilization; green e-commerce; packaging reduction; recycling; biodegradable materials; solid waste reduction
33. Tax liabilities	Tax obligations; compliance with tax laws; compliance with administrative regulations; withholding obligations; tax registration
34. Cooperation with law enforcement	Cooperation with law enforcement; preservation of data in accordance with the law

.

Control variables. Drawing on existing relevant studies, the control variables are borrowed from the studies of Dong Shulan [67] and Feng Feng et al. [66], and the control variables selected in this paper are shown in

Table 1. Variable definitions.

Variable Type	Notation	Name	Description
Variable being explained	$ROA$	Return on total assets	Net profit/TA(total assets) balance
	$ROA1$		Net profit/average balance of TA Average balance of TA = (closing balance of TA + opening balance of TA)/2
	$ROA2$		Net profit/average balance of TA Average balance of TA = (closing balance of TA + closing balance of TA in previous year)/2
Explanatory variables	$ESG$	Comprehensive corporate ESG performance	CSI score /100
	$E$	E Performance	CSI score/100
	$S$	S Performance	CSI score/100
	$G$	G Performance	CSI score/100
	$ESG1$	Corporate ESG overall performance rating	CSI Ratings 1–9
Moderator variable	$MED$	Media attention	The number of times the company was reported on by the media in that year/1000
Intermediary variable	$ZJ$	Platform data governance	LN (platform data governance word frequency summation + 1)
	$MEDIA$	Platform reputation governance	Media bias towards corporate news coverage, J − F coefficient
Control variable	$LEV$	Gearing	Total liabilities/Total assets
	$GRO$	Gross operating income growth rate	(Gross operating income for the current year −amount of gross operating income for the same period of the previous year)/amount of gross operating income for the same period of the previous year
	$AGE$	Age of business	Sample year − firm listing year + 1
	$SHRER1$	Shareholding concentration	Percentage of shares held by the company’s largest outstanding shareholders
	$BD$	Board size	LN (number of board membersi,t)
	$FIX$	Fixed asset ratio	Net fixed assets/TAi,t
	$CASHFLOW$	Cash flow	Operating cash flow/total liabilitiesi,t
	${YEAR}_t$	Annual fixed effects	Control of time factor effects
	$INDUSTRY$	Industry fixed effect	Controlling the impact of industry factors

, with data from the CSMAR database [68,69]. These include debt-to-asset ratio (LEV), revenue growth rate (GRO), business age (AGE), ownership concentration (SHRER1), board size (BD), fixed assets ratio (FIX), and cash flow (CASHFLOW).

3.3. Estimation Method

In order to test Hypothesis H1, this paper constructs the following total effect model (1) to test the impact of ESG performance of platform enterprises on corporate financial performance. Among them, ROA represents the financial performance of the enterprise and ESG represents the ESG performance of the platform enterprise. α₁ is the coefficient of ESG performance of platform enterprises. If α₁ is regular, it indicates that implementing ESG performance of platform enterprises will improve the financial performance of enterprises. The research hypothesis in this paper supports the result that α₁ is positive. Both α₂–α₈ represent the relationship between the relevant control variables and the financial performance of the firm. α₉ is the time fixed effect, α₁₀ is the industry fixed effect, ε is the random interference term.

$$\begin{gathered} {ROA}_{i,t}={\alpha }_0+{\alpha }_1{ESG}_{i,t}+{\alpha }_2{LEV}_{i,t}+{\alpha }_3{GRO}_{i,t}+{\alpha }_4{AGE}_{i,t}+{\alpha }_5{SHRER1}_{i,t}+{\alpha }_6{BD}_{i,t}+{\alpha }_7{FIX}_{i,t}+ \\ {\alpha }_8{CASHFLOW}_{i,t}+\sum {\alpha }_9{YEAR}_{i,t}+\sum {\alpha }_{10}{INDUSTRY}_{i,t}+{\epsilon }_{i,t} \end{gathered}$$

(1)

To test hypothesis H2, the following moderating effect model (2) is constructed to analyze the effect of media attention on the relationship between ESG and financial performance of platform enterprises.

$$\begin{gathered} {ROA}_{i,t}={\beta }_0+{\beta }_1{ESG}_{i,t}+{\beta }_2{MED}_{i,t}+{\beta }_3{ESG}_{i,t}\times {MED}_{i,t}+{\beta }_4{LEV}_{i,t}+{\beta }_5{GRO}_{i,t}+{\beta }_6{AGE}_{i,t} \\ +{\beta }_7{SHRER1}_{i,t}+{\beta }_8{BD}_{i,t}+{\beta }_9{FIX}_{i,t}+{\beta }_{10}{CASHFLOW}_{i,t}+\sum {\beta }_{11}{YEAR}_{i,t} \\ +\sum {\beta }_{12}{INDUSTRY}_{i,t}+{\epsilon }_{i,t} \end{gathered}$$

(2)

In order to test hypothesis H3 and H4, the following intermediary effect models (3), (4), (5) and (6) are constructed to analyze the intermediary effect between platform reputation governance and platform data governance on platform enterprise ESG performance and financial performance, that is, whether platform enterprise ESG performance can further promote the company’s financial performance through platform reputation governance and platform data governance

$$\begin{gathered} {MEDIA}_{i,t}={\gamma }_0+{\gamma }_1{ESG}_{i,t}+{\gamma }_2{LEV}_{i,t}+{\gamma }_3{GRO}_{i,t}+{\gamma }_4{AGE}_{i,t}+{\gamma }_5{SHRER1}_{i,t}+{\gamma }_6{BD}_{i,t}+{\gamma }_7{FIX}_{i,t} \\ +{\gamma }_8{CASHFLOW}_{i,t}+{\sum \gamma }_9{YEAR}_{i,t}+{\sum \gamma }_{10}{INDUSTRY}_{i,t}+{\epsilon }_{i,t} \end{gathered}$$

(3)

$$\begin{gathered} {ROA}_{i,t}={\lambda }_0+{\lambda }_1{ESG}_{i,t}+{\lambda }_2{MEDIA}_{i,t}+{\lambda }_3{LEV}_{i,t}+{\lambda }_4{GRO}_{i,t}+{\lambda }_5{AGE}_{i,t}+{\lambda }_6{SHRER1}_{i,t}+{\lambda }_7{BD}_{i,t} \\ +{\lambda }_8{FIX}_{i,t}+{\lambda }_9{CASHFLOW}_{i,t}+\sum {\lambda }_{10}{YEAR}_{i,t}+\sum {\lambda }_{11}{INDUSTRY}_{i,t}+{\epsilon }_{i,t} \end{gathered}$$

(4)

$$\begin{gathered} {ZJ}_{i,t}={\gamma }_0+{\gamma }_1{ESG}_{i,t}+{\gamma }_2{LEV}_{i,t}+{\gamma }_3{GRO}_{i,t}+{\gamma }_4{AGE}_{i,t}+{\gamma }_5{SHRER1}_{i,t}+{\gamma }_6{BD}_{i,t}+{\gamma }_7{FIX}_{i,t} \\ +{\gamma }_8{CASHFLOW}_{i,t}+{{\sum \gamma }_{9}YEAR}_{i,t}+{{\sum \gamma }_{10}INDUSTRY}_{i,t}+{\epsilon }_{i,t} \end{gathered}$$

(5)

$$\begin{gathered} {ROA}_{i,t}=\lambda 0+{\lambda }_1{ESG}_{i,t}+{\lambda }_2{ZJ}_{i,t}+{\lambda }_3{LEV}_{i,t}+{\lambda }_4{GRO}_{i,t}+{\lambda }_5{AGE}_{i,t}+{\lambda }_6{SHRER1}_{i,t}+{\lambda }_7{BD}_{i,t}+{\lambda }_8{FIX}_{i,t} \\ +{\lambda }_9{CASHFLOW}_{i,t}+\sum {\lambda }_{10}{YEAR}_{i,t}+\sum {\lambda }_{11}{INDUSTRY}_{i,t}+{\epsilon }_{i,t} \end{gathered}$$

(6)

4. Results

4.1. Descriptive Statistics

Table 2. Descriptive statistics.

Variant	Size	Mean Value	Upper Quartile	(Statistics) Standard Deviation	Minimum Value	Maximum Values
ROA	1060	0.025	0.035	0.078	−0.358	0.190
ZJ	1060	3.417	3.466	0.761	1.099	5.663
MEDIA	1060	0.195	0.206	0.211	−0.349	0.644
MED	1060	0.376	0.206	0.515	0.027	3.684
ESG	1060	0.736	0.737	0.052	0.569	0.852
LEV	1060	0.378	0.349	0.192	0.051	0.830
GRO	1060	0.174	0.113	0.424	−0.498	2.673
AGE	1060	13.130	12.000	6.854	2.000	29.000
SHRER1	1060	0.282	0.242	0.141	0.070	0.630
BD	1060	2.096	2.197	0.204	1.609	2.485
FIX	1060	0.099	0.068	0.108	0.001	0.567
CASHFLOW	1060	0.172	0.114	0.297	−0.483	1.527

shows the descriptive statistical results of the main variables in this study. The results show that the minimum value and maximum value of enterprise ESG on the explanatory variable platform are 0.569 and 0.852, respectively, indicating that there is a large difference between the scores of enterprise ESG on different platforms. Other control variables were consistent with existing studies.

4.2. Results of Regression

4.2.1. ESG Master Hypothesis and Sub-Dimensional Tests

From

Table 3. Main regression results.

Variable Sample Type	Full Sample	ROA (E Dimension)	ROA (S Dimension)	ROA (G Dimension)
ESG (E/S/G)	0.245 ***	0.038	0.122 ***	0.182 ***
	(5.077)	(1.456)	(4.268)	(4.739)
LEV	−0.052 ***	−0.053 ***	−0.064 ***	−0.036 **
	(−3.249)	(−3.216)	(−3.797)	(−2.244)
GRO	0.024 ***	0.025 ***	0.025 ***	0.023 ***
	(4.510)	(4.493)	(4.485)	(4.232)
AGE	0.001 *	0.001 **	0.001 **	0.001
	(1.667)	(2.106)	(2.322)	(1.612)
SHRER1	0.062 ***	0.064 ***	0.069 ***	0.051 ***
	(4.596)	(4.614)	(4.917)	(3.820)
BD	0.018	0.015	0.015	0.020
	(1.371)	(1.134)	(1.115)	(1.521)
FIX	−0.065 ***	−0.063 ***	−0.060 ***	−0.078 ***
	(−3.093)	(−2.857)	(−2.807)	(−3.729)
CASHFLOW	0.059 ***	0.063 ***	0.059 ***	0.062 ***
	(7.746)	(8.083)	(7.627)	(8.045)
_cons	−0.191 ***	−0.046	−0.095 ***	−0.174 ***
	(−4.506)	(−1.473)	(−2.712)	(−4.099)
Year Fe	Yes
Industry Fe
N	1060	1060	1060	1060
r2_a	0.233	0.209	0.222	0.231

Notes: * p < 0.1, ** p < 0.05, *** p < 0.01.

, platform enterprises with higher ESG have better financial performance, suggesting that H1 is established. In the dimension regression, it is found that the regression between E performance and financial performance is not significant. S performance and G performance are significantly positively correlated with financial performance. For platform enterprises, the degree of direct environmental pollution is not too high, and the sensitivity of environmental performance to financial performance is relatively low. Social performance and governance performance are, as such, more important for platform enterprises.

4.2.2. Moderating Effects Test

According to

Table 4. Moderating effect.

Variant	ROA	ROA
ESG	0.245 ***	0.191 ***
	(0.048)	(0.061)
LEV	−0.052 ***	−0.058 ***
	(0.016)	(0.016)
GRO	0.024 ***	0.025 ***
	(0.005)	(0.005)
AGE	0.001 *	0.001
	(0.000)	(0.000)
SHRER1	0.062 ***	0.065 ***
	(0.014)	(0.014)
BD	0.018	0.018
	(0.013)	(0.013)
FIX	−0.065 ***	−0.061 ***
	(0.021)	(0.021)
CASHFLOW	0.059 ***	0.055 ***
	(0.008)	(0.008)
MED		−0.111 *
		(0.058)
$\mathrm{ESG}\ast$MED		0.173 **
		(0.081)
_cons	−0.191 ***	−0.155 ***
	(0.042)	(0.048)
N	1060.000	1060.000
r2	0.260	0.267
r2_a	0.233	0.239

Notes: * p < 0.1, ** p < 0.05, *** p < 0.01.

, media attention positively regulates the relationship between the two, and H2 is verified. This may be due to the fact that media attention reduces the information asymmetry between stakeholders and increases their awareness of ESG, thus improving the financial performance of enterprises.

4.2.3. Mediation Effect Test

(1)

Platform reputation governance

According to

Table 5. Regression results for the mediating effect of platform reputation governance.

Variant	ROA	MEDIA	ROA
ESG	0.245 ***	0.786 ***	0.176 ***
	(5.077)	(6.957)	(3.751)
LEV	−0.052 ***	0.038	−0.055 ***
	(−3.249)	(1.015)	(−3.540)
GRO	0.024 ***	0.049 ***	0.020 ***
	(4.510)	(3.239)	(3.812)
AGE	0.001 *	0.000	0.001 *
	(1.667)	(0.272)	(1.658)
SHRER1	0.062 ***	0.084 *	0.055 ***
	(4.596)	(1.882)	(4.225)
BD	0.018	0.080 ***	0.011
	(1.370)	(2.652)	(0.857)
FIX	−0.065 ***	−0.006	−0.065 ***
	(−3.093)	(−0.097)	(−3.197)
CASHFLOW	0.059 ***	0.028	0.057 ***
	(7.746)	(1.262)	(7.596)
MEDIA			0.088 ***
			(6.321)
_cons	−0.191 ***	−0.927 ***	−0.110 ***
	(−4.506)	(−8.432)	(−2.655)
Year Fe	Yes
Industry Fe
N	1060	1060	1060
r2_a	0.233	0.233	0.276

Notes: * p < 0.1, *** p < 0.01.

,The indirect mediating effect of platform reputation governance between ESG and the financial performance of platform enterprises exists.

(2)

Platform data governance

Table 6. Regression results for the mediating effect of platform data governance.

	ROA	ZJ	ROA
ESG	0.245 ***	2.774 ***	0.224 ***
	(5.077)	(7.281)	(4.376)
LEV	−0.052 ***	0.171	−0.053 ***
	(−3.249)	(1.262)	(−3.324)
GRO	0.024 ***	0.111 **	0.024 ***
	(4.510)	(2.168)	(4.381)
AGE	0.001 *	−0.011 ***	0.001 *
	(1.667)	(−3.092)	(1.841)
SHRER1	0.062 ***	−0.583 ***	0.067 ***
	(4.596)	(−3.830)	(4.706)
BD	0.018	0.508 ***	0.014
	(1.370)	(4.935)	(1.088)
FIX	−0.065 ***	−0.693 ***	−0.060 ***
	(−3.0923)	(−3.573)	(−2.837)
CASHFLOW	0.059 ***	−0.000	0.059 ***
	(7.746)	(−0.001)	(7.709)
ZJ			0.007 *
			(1.843)
_cons	−0.191 ***	−0.400	−0.188 ***
	(−4.506)	(−1.125)	(−4.412)
Year Fe	Yes
Industry Fe
N	1060	1060	1060
r2_a	0.233	0.342	0.236

Notes: * p < 0.1, ** p < 0.05, *** p < 0.01.

shows that the indirect mediating effect of platform data governance between ESG and the financial performance of platform enterprises exists. That is to say, the performance of ESG by platform enterprises can affect the financial performance of enterprises through platform data governance, suggesting that H4 is established. Platform enterprises actively implement ESG, which meets the expectations of stakeholders for platform enterprises, enhances the public ‘s confidence in platform enterprises, and affects the financial performance of enterprises through high-level platform data governance.

4.3. Robustness Tests

4.3.1. Replace the Explained Variable

This paper replaces the explanatory variables with ROA1 and ROA2 calculated by different calibers of the return on equity as an indicator of financial performance. In addition, the return on invested capital (ROIC) is used as an explanatory variable for regression. It can be found that the reliability of H1 is proved again (as shown is

Table 7. Robustness tests of replacing the explained variable.

	(1)	(2)	(3)
Variant	ROA1	ROA2	ROIC
ESG	0.240 ***	0.240 ***	0.354 ***
	(5.388)	(5.377)	(4.743)
LEV	−0.049 ***	−0.049 ***	−0.037
	(−3.326)	(−3.362)	(−1.144)
GRO	0.028 ***	0.030 ***	0.034 ***
	(5.110)	(5.547)	(2.668)
AGE	0.000	0.001	0.002 **
	(1.199)	(1.297)	(2.428)
SHRER1	0.055 ***	0.057 ***	0.071 ***
	(4.256)	(4.398)	(3.270)
BD	0.015	0.016	0.020
	(1.286)	(1.322)	(1.130)
FIX	−0.070 ***	−0.070 ***	−0.123 **
	(−3.628)	(−3.648)	(−2.510)
CASHFLOW	0.065 ***	0.065 ***	0.074 ***
	(8.520)	(8.500)	(5.244)
_cons	−0.176 ***	−0.178 ***	−0.232 ***
	(−4.485)	(−4.537)	(−3.440)
Year Fe	Yes
Industry Fe
N	1060	1060	1060
r2_a	0.279	0.284	0.186

Notes: ** p < 0.05, *** p < 0.01.

).

4.3.2. Replace Explanatory Variables

At present, there are many ways to measure ESG, mainly including the 1–100 scoring system and 1–10 scoring system. In this paper, the ESG rating of Huazheng is selected to replace the ESG scoring data of Huazheng, and the same method is used for regression. The results support Hypothesis H1.

4.3.3. Change Sample Period

With the outbreak of the COVID-19 epidemic in 2020, China’s economy was affected, and the relevant data of platform enterprises may also show an unconventional trend. Therefore, this paper excludes the data from 2020 and beyond, and conducts a repeated test on the whole sample from 2013 to 2019, and its key variables are still significant. The specific regression results are shown in

Table 8. Robustness tests of replacing explanatory variables and changing sample period.

Variable Test Method	Substitution of Explanatory Variables	Change Sample Period
ESG1	0.011 ***
	(4.910)
ESG		0.222 ***
		(3.521)
LEV	−0.051 ***	−0.067 ***
	(−3.227)	(−3.031)
GRO	0.024 ***	0.021 ***
	(4.502)	(4.303)
AGE	0.001 *	0.001 **
	(1.685)	(2.324)
SHRER1	0.062 ***	0.050 ***
	(4.595)	(3.153)
BD	0.018	−0.004
	(1.340)	(−0.271)
FIX	−0.065 ***	−0.043
	(−3.043)	(−1.620)
CASHFLOW	0.060 ***	0.053 ***
	(7.823)	(6.352)
_cons	−0.059 *	−0.134 **
	(−1.893)	(−2.413)
Year Fe	Yes
Industry Fe
N	1060	742
r2_a	0.229	0.242

Notes: * p < 0.1, ** p < 0.05, *** p < 0.01.

.

4.3.4. 2SLS

In this paper, 2SLS is used for endogenous test, and the conclusion is basically stable. Please see

Table 9. 2SLS regression results.

Variable Point VA	Phase I	Phase II
	ESG	ROA
L.ESG	0.690 ***
	(29.50)
ESG		0.322 ***
		(5.19)
LEV	0.002	−0.050 ***
	(0.24)	(−3.40)
GRO	−0.002	0.022 ***
	(−0.56)	(3.68)
AGE	0.001 ***	0.001
	(2.84)	(1.33)
SHRER1	−0.016 *	0.068 ***
	(−1.68)	(3.81)
BD	0.001	0.018
	(0.16)	(1.53)
FIX	−0.008	−0.049 *
	(−0.57)	(−1.94)
CASHFLOW	0.014 ***	0.061 ***
	(2.73)	(6.61)
_cons	0.220 ***	−0.248 ***
	(9.26)	(−4.63)
N	954	954
R2	0.532	0.238

Notes: * p < 0.1, *** p < 0.01.

.

5. Further Research

5.1. Platform Enterprise Size

Wenxia Cai et al. find that the ESG performance of larger firms improves financial performance more significantly relative to smaller firms [37]. Yan Weixiang and others also found that scale has an impact on the relationship between the two [36]. The size of platform firms largely reflects market share and consumer stickiness, which determines the financial performance of the firms. Therefore, this paper divides platform enterprises into enterprises of different sizes to further analyze their relationship.

According to the average division of enterprise scale, the α₁ coefficient of small and medium-sized enterprises is smaller than for other enterprises, and the performance of ESG by large-scale platform enterprises has a stronger effect on financial performance.

5.2. Platform Enterprise Ownership Type

According to the different nature of property rights, this paper is divided into two categories to explore the impact of different nature of enterprises. It can be seen that the two types of enterprises are significantly positively correlated. The Fisher combination test is further carried out in this paper, and the results show that the difference between the two coefficients is not significant. Therefore, the nature of property rights results in no significant difference in the relationship between the two.

5.3. Platform Enterprise Classification

In November 2021, the Internet Platform Classification and Rating Guidelines (Draft) was issued by the State Administration for Market Supervision and Administration. Internet platform enterprises were classified into six platform types, namely online sales, life services, social entertainment, financial services, information, and computing applications. This paper combines the above six division bases and classifies the first four types as commercial service platforms and the last two types as network information platforms according to the dominant type of platform users as individual consumers (C2C) or commercial enterprises (C2B).

Table 10. Heterogeneity analysis.

	Broad-Scale	Small- and Medium-Sized	State Enterprise	Non-State Enterprise	Web-Based Information Technology	Business Services
ESG	0.179 ***	0.157 **	0.313 ***	0.223 ***	0.318 ***	0.057
	(2.725)	(1.982)	(3.537)	(3.550)	(5.615)	(0.485)
LEV	−0.076 ***	−0.094 ***	−0.043 **	−0.053 **	−0.068 ***	0.006
	(−3.861)	(−3.019)	(−2.094)	(−2.315)	(−3.905)	(0.147)
GRO	0.017 ***	0.040 ***	0.020 ***	0.025 ***	0.019 ***	0.039 ***
	(3.002)	(4.054)	(2.721)	(3.840)	(3.032)	(3.508)
AGE	0.001	−0.000	0.001	0.002 ***	0.000	0.001
	(1.135)	(−0.366)	(1.030)	(2.753)	(0.827)	(0.901)
SHRER1	0.043 **	0.071 ***	0.104 ***	0.054 ***	0.061 ***	0.068 **
	(2.114)	(3.376)	(4.341)	(2.888)	(3.596)	(2.139)
BD	0.011	0.010	−0.022	0.040 **	0.006	0.039
	(0.732)	(0.499)	(−1.356)	(2.209)	(0.382)	(1.132)
FIX	−0.015	−0.043	−0.006	−0.120 **	−0.099 *	−0.047
	(−0.728)	(−1.348)	(−0.244)	(−2.506)	(−1.921)	(−1.126)
CASHFLOW	0.053 ***	0.052 ***	0.067 ***	0.058 ***	0.052 ***	0.079 ***
	(3.056)	(6.047)	(3.159)	(7.226)	(5.831)	(4.460)
_cons	−0.078	−0.097	−0.207 ***	−0.281 ***	−0.227 ***	−0.155 **
	(−1.052)	(−1.500)	(−2.752)	(−4.255)	(−3.969)	(−1.974)
Year Fe	Yes
Industry Fe
N	499	561	347	713	770	290
r2_a	0.221	0.286	0.374	0.221	0.247	0.210

Notes: * p < 0.1, ** p < 0.05, *** p < 0.01.

shows that the ESG of network information platforms is significantly positively correlated with financial performance, while the ESG of business service platforms is not correlated with financial performance. The possible reason for this is that the users of network information platforms are mostly enterprises, and enterprises that are users of the same ecosystem pay more attention to platform ESG, so the impact is more significant. While the users of business service platforms are mostly individual consumers, the perception of platform ESG is weaker, so the impact is not significant.

6. Discussion and Conclusions

This paper takes China’s Shanghai and Shenzhen A-share listed platform enterprises from 2013 to 2022 as research samples to explore the relationship between ESG and the financial performance of platform enterprises. The study found that (1) the ESG performance of platform enterprises has a positive impact on the financial performance of enterprises, that is, the better the ESG performance of platform enterprises, the better the financial performance. This shows that the platform enterprises fulfill the responsibility of ESG, which is conducive to the establishment of a good corporate image. It not only enhances the confidence of stakeholders in the platform enterprises, but also enhances the trust of consumers in the platform enterprises, which in turn affects ROA. The results show that there is no correlation between E performance and financial performance, while there is a significant positive correlation between S and G performance and financial performance. Therefore, platform enterprises should pay attention to balancing the behavior, performance and effect of each dimension. (2) Media attention will enhance the positive impact of platform enterprise ESG performance on financial performance. (3) Platform reputation governance and platform data governance are two paths for platform enterprise ESG performance to improve financial performance. (4) The nature of property rights has no significant influence on the relationship between the ESG performance and financial performance of platform enterprises, while the scale and type of platform enterprises have heterogeneity in the relationship between ESG performance and financial performance of platform enterprises. Among them, the relationship between ESG and the financial performance of large-scale platform enterprises and network information platform is more significant. Compared with previous scholars who have studied the relationship between corporate social responsibility and corporate value from the perspective of platform corporate social responsibility, this paper studies the relationship between more standardized ESG performance and financial performance, and further helps to achieve sustainable development goals.

This paper provides some new insights into the relationship between ESG and the financial performance of platform enterprises. Based on the findings, this study suggests the following. First, platform companies should actively fulfill their ESG responsibilities to obtain positive media reports and should integrate the concept of ESG into the whole process of production, sales, operations, and other business activities, and actively disclose ESG-related information to alleviate information asymmetry. At present, the scope of influence of platform enterprises is far greater than that of other enterprises. The influence of ESG performance of platform enterprises is deep, the scope is wide, the difficulty of governance is large, and the logic behind the collision is fierce, far beyond the traditional social responsibility issues represented by public welfare and donation. How to make the platform enterprises develop better and more steadily, strengthen compliance management and platform data governance, and deeply understand the connotation of the Internet platform to implement the main responsibility opinion requirements is a problem that each platform enterprise needs to pay attention to. In addition, platform enterprises should further strengthen cooperation and coordination with various stakeholders on ESG of platform enterprises according to their different types of attributes, so as to improve their economic benefits and social influence. Second, the government should formulate and improve relevant policies on ESG to guide platform enterprises to actively implement ESG strategies. On the one hand, enterprises should further implement and improve the platform data governance norms and requirements in the Guidelines for the Implementation of Main Responsibilities of Internet Platforms (Draft for Comments). On the other hand, the government should strengthen the formulation and implementation of relevant policies to effectively guide the development of ESG. By improving the financial support system and review and monitoring system of ESG information, the platform enterprises can be guided to disclose ESG-related information reasonably and legally, which not only helps stakeholders in decision-making, but also helps the platform enterprises to make decisions. It also actively guides the platform enterprises to implement the concept of responsible investment and to actively layout the enterprise ESG strategy. Third, decision-makers should pay attention to the ESG information disclosure status of platform enterprises and media reports of related enterprises, give priority to those enterprises with higher quality of ESG information disclosure and more positive media reports, and reduce information asymmetry in order to provide more financial support. The improvement of the platform’s corporate governance ability helps to empower investors with the value investment concept and value creation process. In addition, investors pay more attention to and distinguish between different types of platform enterprises, which helps them to make more scientific and reasonable investment decisions.

In this paper, the selection of platform enterprises is based on the previous research methods of relevant scholars. However, there is no perfect selection standard for platform enterprises. There is still some subjectivity in the selection of platform enterprises. The platform data governance is selected from the ‘Guidelines for the Implementation of the Main Responsibility of Internet Platforms (Exposure Draft)’ drafted by the State Administration for Market Regulation. Firstly, the keywords were selected, and then the frequency of platform data governance keywords in the annual report of listed companies was counted using text analysis. Finally, the word frequency of all indicators was summed up and logarithmically processed, that is, ln (x + 1). There are two problems in this statistical evaluation. One is that the keywords captured are not typical enough, and the second is that the keywords captured by text analysis are not accurate enough, so the follow-up research needs to be further improved and optimized.