The Impact of Environmental, Social, and Governance Performance on the Total Factor Productivity of Textile Firms: A Meditating-Moderating Model

Abstract

Today’s world is experiencing a great change that has not been seen in a hundred years, with a tense and complex world situation; under the influence of the Israeli–Palestinian conflict, trade friction between China and the U.S., and other events, enterprises need to choose good tactics to achieve strategic development. Environmental, Social, and Governance (ESG) is an indicator that measures the non-financial performance of an enterprise; this article takes listed companies in China’s textile industry from 2015 to 2022 as a research sample and utilizes a bi-directional fixed-effect model that controls for time and individuals to empirically analyze the relationship between ESG performance and corporate total factor productivity (TFP). The results show the following: (1) the better the corporate ESG performance, the higher the TFP; (2) the mechanism test results show that corporate ESG performance promotes TFP by improving green innovation capacity and enhancing corporate human capital, and green innovation and human capital play a partially mediating role; (3) the moderation test shows that agency costs play a weakening role in ESG performance, positively affecting corporate total factor productivity; (4) the heterogeneity analyses found that enterprises are more significantly affected by ESG among non-state-owned enterprises and in the central region. The results of the study provide empirical evidence to guide textile enterprises to actively fulfill ESG performance to enhance enterprise total factor productivity and achieve high quality and sustainable development.

1. Introduction

In the past 20 years, the textile industry has been booming; chemical fiber production and textile raw materials and textile products import and export amounts are showing an overall upward trend. Chemical fiber production reached 66,978,400 tonnes in 2022, which was an increase of 38.6% compared to 2015, where exported textile raw materials and textile products were worth CNY 211,188,787 million in 2022, which, when compared to 2015, constituted an increase of 26.4%. At the same time, the relevant data show that both state-controlled textile enterprises and private textile enterprises are growing in China [1,2,3,4,5,6,7,8]; the above detailed data are shown in Figure 1 and Figure 2 below. The textile industry has been developing rapidly in recent years. Still, the traditional textile industry consumes a lot of energy and is prone to have a bad impact on the environment. As can be seen in Figure 3 below, electricity consumption in China’s textile industry has remained stable year by year, while at the same time, coal consumption has shown a significant downward trend, and natural gas consumption as a clean energy source has shown a significant upward trend, which reflects China’s emphasis on energy conservation and emission reduction in the traditional textile industry, and the pursuit of the concept of green and sustainable development. Due to the impact of the COVID-19 epidemic outbreak in early 2020, the total energy consumption plummeted in 2020, and after the economy slowly recovered, the total energy consumption showed a small peak in 2021.

China promulgated the New Environmental Protection Law in 2015 [9], and in 2023, the China Textile Industry Federation pointed out through the Environmental, Social, and Governance (ESG) High Standards Summit Forum that ESG disclosure work in the textile industry can enhance the industry’s corporate management level, improve sustainable development [10], and provide a reference path for high-quality development. ESG is a measure of non-financial performance indicators of enterprises [11] and was first proposed by the United Nations Global Compact in 2004. In recent years, sustainability and green development have attracted much attention, and there are numerous topics about ESG performance. Qiang Sun argues that employees’ perceptions of in-house ESG performance are conducive to green innovation [12], Yongjie Zhu argues that ESG performance improvement is affected by banks’ digital transformation [13], and Xiaochen Yu believes that the ESG advantage of enterprises is to improve the total factor productivity of enterprises by increasing internal labor costs and reducing agency costs [14], especially applicable to economically underdeveloped regions, patent-intensive industries, and regions with low environmental risks [15]. Meanwhile, social performance in ESG can act as a mediator through digital transformation to promote corporate ESG performance and thus corporate Total Factor Productivity (TFP) [16]. The importance of ESG in enterprises is gradually highlighted, and it can be applied to various business processes of enterprises, such as research and development(R&D) investment [17], financial performance [18], corporate finance [19], stock returns [20], cost of debt [21], corporate performance [22], and so on.

The “14th Five-Year Plan” clearly puts forward green development; the 20th National Congress report points out that we should realize high-quality sustainable development, and that improving TFP is an important connotation of it [23]. TFP is the main driving force for the technological and structural upgrading of enterprises, and it is also an important prerequisite for the sustainable development of enterprises [24]. Scholars have examined the factors affecting TFP from both the internal factors of enterprises and the external environment. Shaohui Zou believes that the digital economy enhances total factor productivity through the upgrading of urban industrial structure and urban innovation and entrepreneurship [25], and Boqiang Lin believes that venture capital and government subsidies help to increase the total factor productivity of renewable energy enterprises [26,27], and Shasha Jin argues that digital finance has a positive effect on TFP [28]; in addition to this, total factor productivity is also affected by financial crises, corporate tax [29], the misallocation of resources [30], and career change and financial deepening [31].

However, what is the association between ESG indicators and TFP in textile companies? This is a subject that has been hardly studied by only a few scholars. In this context, the study in this paper has four theoretical implications. First, using listed Chinese textile firms as the study sample, the two-way fixed-effects model strongly affirms theoretically the important role of ESG performance in improving firms’ total factor productivity. This line of analysis expands the path for textile firms to improve total factor productivity. Second, regarding the limited research on the influence of total factor productivity from the internal factors of enterprises, our article profoundly analyses the mechanism of green innovation and the human capital of enterprises affecting the total factor productivity of textile enterprises from a theoretical point of view and comprehensively describes the specific relationship between ESG performance, green innovation, and total factor productivity, as well as between ESG performance, human capital, and total factor productivity. Furthermore, it is empirically verified that agency cost plays a weakening effect in the process of ESG performance to improve the total factor productivity of enterprises, which strengthens our attention to the agency cost of textile enterprises. Finally, the existing theoretical results are expanded by analyzing whether there is any difference in the impact of firms’ ESG performance on total factor productivity in terms of the nature of the firms and their geographical locations.

The research in this paper has two practical implications. Firstly, for the government, this paper has theoretically confirmed the role of ESG performance on the total factor productivity of enterprises; at this time, the government should strictly improve the relevant ESG policies and standardize the content, rules, and procedures of ESG disclosure for enterprises, as well as ensure the accuracy of the relevant ESG information released to the society to avoid harming the interests of investors. Secondly, for enterprises, this paper supports the promotion of the role of green innovation and human capital mechanisms; textile enterprises should increase investment in research and development and the use of technological innovation to achieve the transformation and upgrade of the textile industry through intelligent, mechanized production creativity to improve the total factor productivity of enterprises. At the same time, it is necessary to give full play to the advantages of human capital, recruit modern advanced talents, and operate intelligently.

2. Theoretical Analysis and Assumptions of the Study

TFP, as an important indicator in the field of economics, is a variable that favors economic growth. Enhancing the total factor productivity of textile enterprises requires technological upgrades and changes in the management mode of textile enterprises, thus promoting the high-quality development of the textile industry.

ESG is considered a matter of high importance by scholars all over the world, both in Europe and in Asia; in recent years, there have been more and more economic studies related to ESG topics. For the relationship between the European banking sector and financial stability, scholars use different experimental methods to express their views: Toth et al. used a panel regression method to analyze 243 European banks from 2002 to 2018, and concluded that ESG performance not only reduces non-performing loans, but also positively affects regulatory capital, ultimately confirming that ESG performance is beneficial to the financial stability of the banking sector [32]. Laura Chiaramonte, on the other hand, used a double-difference approach to derive that in times of financial turbulence, the longer the European banking sector discloses its ESG information, the greater the financial stability [33]. Iulia Lupu used quantitative analysis to verify that European banks’ ESG scores affect their financial stability and that this effect is non-linear [34]. For the relationship between ESG performance and firms’ financial performance, Caterina De Lucia used European-listed firms’ data and analyzed them using machine learning and logistic regression models to discover that ESG performance can bring better financial performance for listed firms; they also proposed to focus on environmental innovation, employment productivity, and diversity in ESG [35]. Phoebe Koundouri, based on a study sample of the top 50 European firms in Europe in various industries in terms of ESG performance, concluded that firms with better ESG performance have a relatively lower equity risk; although this conclusion excludes firms in the automotive industry, for any industry, the better the ESG performance, the better the firm’s profitability and financial performance [36]. De Franco C concluded that ESG controversies can hurt stock returns in Europe and the US, whereas for the Asia–Pacific stock market, the stock market did not hurt ESG controversies or declining ESG ratings [37].

2.1. ESG, Green Innovation, and Total Factor Productivity

Green innovation does not have to be aimed at reducing the environmental burden but can produce significant environmental benefits [38], and it has been shown that firms with good ESG performance can significantly improve the level of green innovation in their firms [39]. Green innovation, on the other hand, can promote the total factor productivity of enterprises in five aspects: absorbing new technologies [40], developing invention patents [41], alleviating financing constraints [42], providing R&D funding sources, and improving management short-sightedness [42]. For traditional textile enterprises with high pollution levels and high resource consumption, green innovation utilizing new technological achievements cannot only reduce the dependence on the original polluting production methods but can also enhance production efficiency. Secondly, green innovation is more difficult to research and develop, with more R&D investment, a longer R&D cycle, greater R&D risk, and a greater need for financial support than ordinary innovation. According to the information asymmetry theory, ESG information disclosure by enterprises is conducive to investors’ clearer understanding of the enterprise’s operation [43], easing financing resistance. Finally, management short-sightedness refers to the behavior of management that is too conservative and resistant to green innovation. According to the agency theory, companies with good ESG performance will improve the operational efficiency of the enterprise by restraining the management, reduce the short-sighted behavior of the management [44], thereby accelerating the development of green innovation strategies by the management to improve the green innovation capability. Based on the above analysis, the following hypotheses are proposed:

H1. 

ESG indicators are positively correlated with the TFP of textile enterprises.

H2. 

ESG performance improves corporate TFP by improving green innovation capabilities.

2.2. ESG, Human Capital, and Total Factor Productivity

Human capital is innovative and creative, and one of the factors that enhance economic indicators. A wise investment in it can improve the productivity and innovation ability of enterprises [45], and the innovation ability of enterprises is conducive to the formation of the unique core competitive advantage of the enterprise, which enables the enterprise to achieve the purpose of long-term survival and development [46]. According to the reputation theory, companies with good ESG performance can improve their reputation, not only by expanding their brand value [47], but also by attracting high-quality talents and improving the efficiency of their human capital [48], and by learning from external excellent experience and management methods, high-level human capital will bring greater value creation efficiency to enterprises. A greater value creation efficiency will improve the overall TFP of the enterprise. Based on this, this paper proposes a third hypothesis:

H3. 

ESG performance improves firms’ total factor productivity by increasing the level of human capital.

2.3. ESG, Agency Cost, and Total Factor Productivity

Agency theory [49] consists of two aspects: the first is the relationship between owners and managers and the second is the separation of ownership and operation [50]. Owners give the right to run the company to managers, and due to the inconsistency of their goals, operators usually prioritize their interests and ignore the interests of owners, which results in agency costs [51]. Whereas ESG disclosure provides new non-financial information, firms with good ESG performance represent good corporate governance, and effective corporate governance reduces conflicts between owners and managers and lowers agency costs [52]. Prior studies have shown that agency costs have a dampening effect on a firms’ total factor productivity growth [53], and high agency costs can bring about information asymmetry problems [54], leading to the loss of capital and talent from firms to the detriment of TFP improvement. In this regard, the fourth hypothesis is proposed:

H4. 

Agency cost negatively moderates the positive impact of ESG performance on corporate total factor productivity.

Figure 4 shows the research model and summarizes the above assumptions.

3. Structure of the Research

3.1. Choosing Specimens and Calculating Data Volume

This paper selects A-share textile enterprise-quoted companies in China from 2015 to 2022 as the research sample; the selected enterprises are based on the Guidelines for Industry Classification of Listed Companies (revised in 2012) issued by the China Securities Regulatory Commission (CSRC) as the standard, and the ESG data are matched with the enterprise total factor productivity data and other data to form the panel data from 2015 to 2022. Meanwhile, the resulting data are processed as follows: (1) ST and ∗ST enterprises are deleted; (2) the continuous variables are reduced by 1% and 99% quantile by year.

3.2. Meaning of the Symbols

3.2.1. Explanatory Variable

Total Factor Productivity (TFP) of enterprises. TFP is a measure of the high-quality development of enterprises, and the current literature will use the Labor-based Leontief Method (LP) method and Ordinary Least Squares (OLS) method to measure the TFP of enterprises. This paper mainly refers to the relevant research of James Levinsohn [55], who chose to use the LP method to calculate total factor productivity in the benchmark regression. The LP method uses intermediate goods inputs to replace the investment amount as a proxy variable, and the intermediate goods inputs are the enterprise’s cost of doing business plus the selling expenses, financial expenses, and administrative expenses minus the enterprise’s depreciation and amortization charged in the current period as well as the payments made for and to the employees. The advantage of replacing the investment amount with the intermediate goods input is that the enterprises with non-positive investment amounts can also be used as the reference sample, which not only can expand the research sample but also can obtain more reasonable empirical results. The formula is shown in Equation (1).

$${TFP-LP}_{it}={\alpha }_0+{\alpha }_1{\mathrm{I}\mathrm{n}K}_{it}+{\alpha }_2{\mathrm{I}\mathrm{n}L}_{it}+{\alpha }_3{\mathrm{I}nM}_{it}+{\epsilon }_{it}$$

(1)

where InK is the logarithm of fixed assets, InL is the logarithm of employees, and InM is the logarithm of cash paid for goods and services.

3.2.2. ESG Performance (ESG)

ESG performance (ESG): The CSI index divides ESG indicators into nine levels, with AAA scoring 9, AA scoring 8, A scoring 7, BBB scoring 6, BB scoring 5, B scoring 4, CCC scoring 3, CC scoring 2, and C scoring 1. The higher the ESG score, the better the ESG performance of the company.

3.2.3. Mediating Variables

To study the mediating path of corporate ESG indicators affecting corporate TFP, this paper sets two mediating variables, which are the Green Indicators (Igi) and the Human Indicators (Human). The specific definitions are as follows:

Enterprises’ green innovation level (Igi): This paper draws on the research of XI Longsheng and adopts the logarithm of the number of green patent applications plus one as a measure of enterprise green innovation [56].

Human capital level (Human): This paper draws on Yao Ruizhen’s study and uses the ratio of bachelor’s degree and above to total employees to measure the human capital level of enterprises [57].

3.2.4. Moderator Variable

Agency cost (AC). Drawing on Tong Yong’s research, this paper uses the first type of agency cost as a measure, which is the ratio of administrative expenses to main business income [58].

3.2.5. Control Variable

The article selected four control variables: the corporate gearing ratio (Lev), corporate growth (Growth), fixed assets (Fixed), and the management shareholding ratio (mshare). Year (Year) and individual (stkcd) fixed effects are also controlled. See

Table 1. Explanatory notes related to variables.

Type	Title	Symbol	Definitions	Source of Data
Explained variables	TFP of enterprises	TFP_LP	LP method for the calculation of TFP of firms	The index constructed by James Levinsohn et al.
Core explanatory variable	ESG performance	ESG	China Securities ESG Rating	WIND database
Mediation variables	Green innovation	Lgi	In (number of green patents filed +1)	WIND database
Mediation variables	Human capital	Human	Total number of persons educated to bachelor or above/total number employed
Adjust variables	Agency costs	AC	Administrative expenses/main business income	CSMAR database
Control variables	Debt-to-asset ratio	Lev	Total liabilities/total equity	WIND database
	Enterprise growth	Growth	Principal operating income grow
	Proportion of fixed assets	Fixed	Tangible/total assets
	Management shareholding	M share	Share ownership of directors /aggregate outstanding equity shares

for specific definitions.

3.3. Modeling

To test the impact of firms’ ESG indicators on their TFP, we follow Xiang Deng et al. [59] to construct the following econometric model:

$$\begin{gathered} {TFP-LP}_{it}={\beta }_0+{\beta }_1{ESG}_{it}+{\beta }_2{Lev}_{it}+{\beta }_3{Growth}_{it}+{\beta }_4{Fixed}_{it}+{\beta }_5{Mshare}_{it} \\ +{\mu }_i+{\lambda }_t+{\epsilon }_{it} \end{gathered}$$

(2)

$$\begin{gathered} {Med}_{it}={\gamma }_0+{\gamma }_1{ESG}_{it}+{\gamma }_2{Lev}_{it}+{\gamma }_3{Growth}_{it}+{\gamma }_4{Fixed}_{it}+{\gamma }_5{Mshare}_{it}+{\mu }_i \\ +{\lambda }_t+{\epsilon }_{it} \end{gathered}$$

(3)

$$\begin{gathered} {TFP-LP}_{it}={\theta }_0+{\theta }_1{ESG}_{it}+{{\theta }_2{Mediam}_{it}+\theta }_3{Lev}_{it}+{\theta }_4{Growth}_{it} \\ +{\theta }_5{Fixed}_{it}+{\theta }_6{Mshare}_{it}+{\mu }_i+{\lambda }_t+{\epsilon }_{it} \end{gathered}$$

(4)

$$\begin{gathered} {TFP-LP}_{it}={\eta }_0+{\eta }_1{ESG}_{it}+{\eta }_2{AC}_{it}+{\eta }_3{ESG}_{it}\ast {AC}_{it}+{\eta }_4{Lev}_{it}+{\eta }_5{Growth}_{it} \\ +{\eta }_6{Fixed}_{it}+{\eta }_7{Mshare}_{it+}{\mu }_i+{\lambda }_t+{\epsilon }_{it} \end{gathered}$$

(5)

In the above model, i denotes an individual firm, and t represents a specific year, controlling for individual fixed effects (μ) and year-fixed effects (λ). Lev, Growth, Fixed, and Mshare are the control variables. Med is a mediator variable, which stands for the level of green innovation and the level of the firm’s human capital.

H1 will be verified by model (2), in which the explanatory variable TFP-LP denotes the total factor productivity of the enterprise, and the explanatory variable ESG denotes the ESG performance of the textile enterprise; if the coefficient of β₁ is positive, it indicates that the ESG performance of the textile enterprise is positively correlated with its total factor productivity, i.e., the better the ESG performance, the higher the total factor productivity of the textile enterprise.

The mediation effect test is needed to verify the mechanism path of corporate ESG indicators on its total factor productivity level. The article adopts Wen Zhonglin’s three-step method [60], using a stepwise regression method for testing. The test steps are divided into three steps: the first step is to regress the explanatory variable ESG on the explanatory variable TFP-LP, i.e., model (2); the second step is to regress the explanatory variable ESG on the mediator variables Lgi and Human, respectively, i.e., model (3); and the third step is to add the mediator variables in the regression of the explanatory variable ESG on the explanatory variable TFP_LP to regress the mediator variables, i.e., model (4).

To test whether corporate agency costs affect the degree of influence of ESG indicators on corporate TFP, the article draws on the study of Shu Huan et al. [61] and establishes a moderated model (5) by taking corporate TFP as the explanatory variable, ESG indicators as the explanatory variable, and agency costs, AC, as the moderating variable.

4. Empirical Results and Analyses

4.1. Description of Relevant Variables

The maximum value of ESG performance is 7, the minimum value is 1, and the mean is 3.902, which side-by-side reflect that the ESG index of the selected enterprises stays between B-BB, and there is still much room for improvement in the ESG performance of textile enterprises. The highest value of enterprise TFP is 10.341 and 6.668 is its lowest value, with a standard deviation of 0.813 and a mean of 8.431, indicating that there is a large gap between the respective TFPs of the selected enterprises. Still, the whole is above a reasonable level. The optimal value of green innovation level is 6.515, the lower limit value is 0, and the standard deviation is 1.581, indicating that the sample firms have a large difference in their respective levels of green innovation, which may be related to the level of education of their employees and corporate strategy. The collation results of the remaining variables are all in the standard range. The data of the relevant variables are shown in

Table 2. Descriptive statistics.

Variable	Observing	Average	Standard Error	Min	Max
TFP_LP	831	8.431	0.813	6.668	10.341
ESG	955	3.902	1.160	1.000	7.000
lev	981	0.374	0.179	0.065	0.908
growth	933	0.160	0.607	−0.660	4.519
fixed	981	0.179	0.129	0.001	0.546
Share	959	0.149	0.215	0.000	0.780
Lgi	981	2.353	1.581	0.000	6.515
human	976	14.897	15.445	0.000	79.271
AC	981	0.077	0.073	0.008	1.007
SOE	967	0.178	0.383	0.000	1.000
Mid	964	0.057	0.232	0.000	1.000

below.

4.2. Baseline Regression Analysis

In this paper, a two-way fixed-effects model controlling for time and individuals is chosen to empirically validate the effect of ESG indicators on firms’ TFP, and the regression analysis is shown in

Table 3. Baseline regression results.

	(1)	(2)	(3)	(4)
	TFP(LP)	TFP(LP)	TFP(LP)	TFP(LP)
ESG	0.046 ***	0.036 **	0.034 **	0.037 **
	(2.610)	(2.122)	(1.965)	(2.452)
lev				1.221 ***
				(9.533)
growth				0.133 ***
				(6.256)
fixed				−0.973 ***
				(−4.287)
mshare				0.626 ***
				(2.706)
_cons	8.165 ***	7.871 ***	7.979 ***	7.588 ***
	(84.960)	(79.655)	(103.436)	(78.707)
Year	No	Yes	Yes	Yes
Stkcd	No	No	Yes	Yes
N	829	829	829	813
R2			0.172	0.333
Adj. R2			0.007	0.193

*** p < 0.01, ** p < 0.05.

. In column (1), without adding control variables and without controlling for time and individual effects, the results show that ESG performance and TFP are significant at the 1% significance level; in column (2), with the addition of controlling for time fixed effects, ESG performance and TFP are still significant; in column (3), with the addition of individual fixed effects, the results are still significant; and in column (4), based on the above, with the addition of the introducing control variables, ESG performance and TFP are significant at the 5% level of significance. This proves the role of ESG performance in enhancing firms’ TFP and validates H1.

4.3. Mediation Effect Test

In Section 2, this paper explains that corporate ESG performance enhances corporate total factor productivity through green innovation and human capital mechanisms, and to validate this mechanism, the mediation effect regression model is used to explain the relationship between the relevant variables based on the benchmark regression.

Table 4. Regression results for mediation effects.

	(1)	(2)	(3)	(4)	(5)
	TFP(LP)	Lgi	TFP(LP)	Human	TFP(LP)
ESG	0.037 **	0.178 ***	0.031 **	0.561 *	0.031 **
	(2.452)	(4.309)	(2.032)	(1.814)	(2.092)
Lgi			0.040 ***
			(2.924)
Human					0.009 ***
					(4.773)
lev	1.221 ***	1.563 ***	1.137 ***	−0.276	1.225 ***
	(9.533)	(4.600)	(8.699)	(−0.108)	(9.718)
growth	0.133 ***	−0.012	0.135 ***	0.404	0.124 ***
	(6.256)	(−0.210)	(6.363)	(0.956)	(5.902)
fixed	−0.973 ***	−0.743	−0.931 ***	−17.196 ***	−0.918 ***
	(−4.287)	(−1.203)	(−4.116)	(−3.714)	(−4.105)
mshare	0.626 ***	0.926 *	0.613 ***	−1.863	0.602 ***
	(2.706)	(1.756)	(2.663)	(−0.471)	(2.641)
_cons	7.588 ***	0.606 **	7.564 ***	12.077 ***	7.509 ***
	(78.707)	(2.314)	(78.603)	(6.159)	(77.946)
Year	do	do	do	do	do
Stkcd	do	do	do	do	do
N	813	910	813	910	813
R2	0.333	0.139	0.342	0.183	0.355
Adj. R2	0.193	−0.038	0.202	0.015	0.219
Sobel-Z		2.287		1.945
intermediary effect-ab		0.006181		0.005799
direct effect		0.030631		0.031013
overall effect		0.036812		0.036812
Percentage of intermediary effects		16.79%		15.75%

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

shows the results of ESG performance affecting corporate TFP mechanism conduction; from the results in column (2), it can be seen that there is a positive correlation between the level of corporate ESG and the level of corporate green innovation, and it is significant at the 1% significance level, with a correlation coefficient of 0.178. This shows that corporate ESG governance will enhance the level of green innovation [53]. Column (3) is the result of adding the level of corporate ESG and the level of corporate green innovation to model (4) at the same time; it can be seen that the regression coefficients of ESG performance and green innovation are 0.031 and 0.04, respectively, which indicates that there is a partial mediating effect of green innovation on the mechanism of corporate ESG performance on TFP, i.e., ESG performance enhances corporate TFP by improving green innovation capability. The coefficient of ESG performance in column (4) is 0.561, which indicates that ESG performance can attract high-level talents to join the enterprise to a certain extent, which in turn improves the level of human capital. Column (5) adds ESG performance and human capital into the model (4) at the same time; it can be seen that ESG is significant at the 5% level, and Human is significant at the 1% level, which indicates that there is a partially intermediary effect on the level of human capital in the mechanism of the enterprise’s ESG performance on the impact of TFP, i.e., ESG performance can enhance corporate TFP by improving green innovation capability. ESG performance can contribute to the improvement of corporate TFP by enhancing the level of human capital.

To be cautious, this paper draws on the theoretical references of Özdil et al. [62] and uses the Sobel test to re-verify the verification of the presence of mediating effects. The specific framework flow chart is shown in Figure 5 below. x is ESG, y is TFP-LP, and m is the mediating variables green innovation (Lgi) and human capital level (Human). c is the regression coefficient β₁ in the model (2), which is the effect of ESG on the TFP regression; a is the regression coefficient γ₁ in the model (3), i.e., the effect of ESG on the mediating variables green innovation (Lgi) and human capital level (Human); b is the effect of the mediating variables Lgi and Human on TFP after controlling for the effect of ESG. c′ is θ1 in the model (4), the direct effect of ESG on TFP after controlling for the effects of the mediating variables Lgi and Human. The product of Sobel’s test ab is the indirect effect, and the sum of the indirect and direct effects is the total effect.

The Sobel test Z value of green innovation in this paper is 2.287, which is greater than 1.96 and is significant at a 5% level; the ab indirect effect is 0.006181, the direct effect is 0.036812, the direct effect and the profile effect have the same positive sign, and there is a part of the mediation effect, which is 16.79%. The Sobel test verifies H2. The Sobel test Z value for human capital is 1.945, which is greater than 1.65 and significant at a 10% level; the ab indirect effect is 0.005799, the direct effect is 0.031013, the direct effect and the indirect effect are both positive, and there is a partial mediating effect, accounting for 15.75%, which verifies H3.

4.4. Reconciliation Analysis of Agency Costs

To study whether there are significant differences on the impact of ESG indicators on firms’ TFP under different agency costs, this paper regresses the relationship between ESG performance, the interaction terms between ESG performance and agency costs, and firms’ total factor productivity, as shown in

Table 5. Regression results of agency cost adjustment mechanism.

	(1)	(2)
	TFP_LP	TFP_LP
ESG	0.028 **	0.099 ***
	(2.060)	(4.724)
AC	−4.873 ***	−2.072 ***
	(−12.903)	(−2.807)
ESG × AC		−0.916***
		(−4.394)
lev	1.170 ***	1.146 ***
	(10.192)	(10.104)
growth	0.092 ***	0.097 ***
	(4.738)	(5.069)
fixed	−0.448 **	−0.439 **
	(−2.159)	(−2.145)
mshare	0.434 **	0.476 **
	(2.088)	(2.319)
_cons	8.038 ***	7.822 ***
	(86.314)	(75.034)
Year	Yes	Yes
Stkcd	Yes	Yes
N	813	813
R2	0.466	0.481
Adj. R2	0.353	0.370

*** p < 0.01, ** p < 0.05.

below. As can be seen from column (1) in the table, the regression coefficient between agency cost and total factor productivity is −4.873, and it is negative at the 1% confidence level, indicating that agency cost hurts the total factor productivity of enterprises. As can be seen from column (2) in the table, the regression coefficient of ESG performance is 0.099 and is significant at the 1% confidence level after adding the moderating variable AC and its interaction with the explanatory variable ESG × AC, indicating that the positive impact of ESG performance on the total factor productivity of enterprises remains significant. Further analysis of the significance level of the moderating variable AC shows that the regression coefficient of ESG × AC is −0.916, and it is significantly negative at the 1% confidence level, indicating that the positive effect of agency cost on ESG performance on enterprise total factor productivity has a weakening effect. The ESG performance of textile enterprises with low agency costs has a greater impact on the total factor productivity of enterprises, which verifies H4.

4.5. Robustness Check

4.5.1. Replacement of Measures of Explanatory Variables

Based on the regression analysis in the previous paper, this paper refers to Xuan Min’s approach and adopts the Business Way Rong Green ESG rating as the core explanatory variable instead of the Huazheng ESG rating in the regression [63], named ESG1. The results are shown in

Table 6. Regression results with replacement of explanatory variables.

	(1)	(2)
	TFP_LP	TFP_LP
ESG1	0.057 ***	0.054 ***
	(6.61)	(6.92)
lev		1.248 ***
		(9.73)
growth		0.138 ***
		(6.13)
fixed		−1.388 ***
		(−6.44)
mshare		−0.059
		(−0.32)
Constant	8.274 ***	8.077 ***
	(125.46)	(91.43)
Observations	831	814
Number of stkcd	130	130
Control	No	Yes
Company FE	Yes	Yes
Year FE	Yes	Yes

*** p < 0.01.

. From the table, it can be seen that the regression coefficient of the second column is significantly positive at the 1% level, which again further verifies hypothesis H1, that is, corporate ESG performance can promote TFP, and the better the ESG performance, the greater that enhancement effect is, indicating that the conclusions of this paper are robust.

4.5.2. Endogeneity Test

Under panel data, this paper selects a two-way fixed-effects model to empirically verify the impact of ESG indicators on corporate TFP after the Hausman test. The p-value of the Hausman test is 0.0189, which negates the original hypothesis that all explanatory variables are exogenous and thus there is an endogeneity problem. To alleviate the endogeneity problem, this paper refers to Wu Peng’s approach [64], choosing the ESG mean value of the same city in the same year as an instrumental variable (ESG-IV), and the regression results are shown in

Table 7. Endogeneity tests.

	(1)	(2)
	First	Second
	ESG	TFP_LP
ESGIV	0.9364 ***
	(8.05)
ESG		0.1499 ***
		(5.24)
lev	0.1891	1.2862 ***
	(0.60)	(8.53)
growth	0.0086	0.1867 ***
	(0.16)	(4.13)
fixed	−0.2065	−1.4362 ***
	(−0.37)	(−7.20)
mshare	1.0102 *	−0.4529 ***
	(1.78)	−3.47)
Constant	1.3181 **	7.3759 ***
	(2.30)	(49.29)
Year	Yes	Yes
Stked	Yes	Yes
Observations	813	813
R-squared	0.630	0.188

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

. Column (1) is the regression result of the first stage; the coefficient of ESG-IV is positive at a 1% level, which indicates that the instrumental variable selected in this paper is highly correlated with ESG, and also shows that the instrumental variable selected in this paper is not a weak instrumental variable. Column (2) is the regression result of the second stage, and the estimation of the second stage shows that the ESG performance of textile enterprises is still significantly positively correlated with the TFP under consideration of the instrumental variable. It is still significantly positively correlated with TFP when the instrumental variables are taken into account, which verifies the reliability of the benchmark regression in the previous section.

5. Heterogeneity Analysis

5.1. Analysis of Property Rights Heterogeneity

Drawing on Zhang Weiying and other scholars [65], this paper categorizes the sample textile firms into state-owned textile firms and non-state-owned textile firms based on the form of ownership. Columns (1) and (2) in

Table 8. Results of the heterogeneity test.

	(1)	(2)	(3)	(4)	(5)
	TFP_LP Government-Owned	TFP_LP Privately Owned	TFP_LP East	TFP_LP Central	TFP_LP West
ESG	−0.034	0.039 **	0.021	0.242 **	0.029
	(−1.182)	(2.527)	(1.305)	(2.509)	(1.016)
lev	−0.170	0.955 ***	1.044 ***	2.230 **	0.651
	(−0.563)	(7.037)	(7.261)	(2.596)	(1.478)
growth	0.185 ***	0.101 ***	0.115 ***	0.246 **	0.252 ***
	(4.131)	(4.694)	(4.988)	(2.658)	(2.833)
fixed	−1.599 ***	−0.439 *	−0.276	−2.360 **	−1.057 *
	(−4.414)	(−1.710)	(−1.096)	(−2.142)	(−1.685)
mshare	6.205	1.609	1.538	−13.845	0.301
	(3.343)	(9.041)	(8.148)	(−0.202)	(2.798)
_cons	8.993 ***	7.480 ***	7.583 ***	6.570 ***	8.003 ***
	(46.311)	(72.871)	(74.925)	(12.008)	(29.118)
Year	Yes	Yes	Yes	Yes	Yes
Stkcd	Yes	Yes	Yes	Yes	Yes
N	141	672	684	49	80
R2	0.455	0.312	0.302	0.546	0.372
Adj. R2	0.266	0.156	0.150	0.248	0.081

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

give the group regression results; the coefficients of ESG are significantly positively correlated at the 5% significance level in non-state-owned enterprises, and are not significant in state-owned enterprises. This occurs because, compared with state-owned enterprises, non-state-owned enterprises have more autonomy, and the better the efficiency of the investment [58], the greater the enhancement of the ESG indicators on the performance of the enterprise, so the ESG performance has an enhancing effect on the TFP of non-state-owned enterprises.

5.2. Analysis of Regional Heterogeneity

This paper draws on the practice of Ye Maosheng and other scholars [66] combined with the National Bureau of Statistics (NBS), the eastern, central, western, and northeastern regions are specifically divided as follows: the 10 eastern provinces (municipalities) include Beijing, Tianjin, Hebei, Shanghai, Jiangsu, Zhejiang, Fujian, Shandong, Guangdong and Hainan; the 6 central provinces include Shanxi, Anhui, Jiangxi, Henan, Hubei, and Hunan; the 12 western provinces (districts and municipalities) include Inner Mongolia, Guangxi, Chongqing, Sichuan, Guizhou, Yunnan, Tibet, Shaanxi, Gansu, Qinghai, Ningxia, and Xinjiang; and the three northeastern provinces include Liaoning, Jilin and Heilongjiang. To avoid the differences in economic policies caused by the differences in policies of one country and two systems, and to avoid the influence of special values on the overall conclusions, this paper temporarily excludes the data related to Taiwan Province of China, Hong Kong Special Administrative Region, and Macao Special Administrative Region. From the regression results in Table 8, it can be seen that there is a positive correlation between ESG and TFP in the central economic zone, which is significant at the 5% level, and not significant in the eastern and western economic zones, which occurs because the central part of the country has begun to undertake the transfer of textile enterprises from the east since as early as 2006. Sun Huaibin, director of the China Textile News Centre, said that the growth rate of investment in the textile industry in central China is much higher than that in the east and west, and the signs of the textile industry migrating from the east to the center are becoming more and more obvious; not only that, but according to the China Textile and Clothing Industrial Parks Development Report 2019, there are about 160 parks in central China, while there are only about 140 parks in the east and 100 parks in the west, and the central parks are focusing on the development of textile intelligent manufacturing, with intelligent grip; these locations are better developed compared to the east and the west, so the ESG performance of the central region has a greater role in boosting enterprises, which in turn improves the TFP of textile enterprises in the central region.

6. Conclusions, Implications, and Limitations of the Study

6.1. Conclusions

Due to the trade friction between China and the United States, the world is experiencing a great change that has not been seen in a hundred years. Furthermore, China’s traditional textile industry has a large amount of energy consumption, and China has put forward the goal of “carbon peak, carbon neutral”; the pursuit of sustainable green development in the context of this paper, with the help of the sample of Chinese textile listed companies in the period of 2015–2022 as a research sample, has offered an in-depth discussion of the relationship and mechanism between ESG performance and TFP of textile companies. This paper explores the relationship and mechanism between ESG performance and the TFP of textile companies. The research conclusions show that (1) ESG indicators can enhance the TFP of textile enterprises, and the conclusions remain valid after the endogeneity test and the robustness test. (2) The test of the mediation effect model proves that in the process of ESG performance affecting total factor productivity, there exists a transmission path from ESG performance to green innovation and then to enterprise total factor productivity, as well as from ESG performance to the level of human capital and then to enterprise total factor productivity. ESG performance can enhance the level of green innovation of the enterprise [67] and also positively affect the level of human capital [48]; the level of green innovation and human capital levels have an enhancing effect on the total factor productivity of textile firms [68]. (3) The moderating effect found that agency cost plays a weakening effect in the process of corporate ESG performance, affecting total factor productivity [69]; i.e., corporate ESG responsibility fulfillment is more likely to obtain TFP enhancement when agency cost is low. (4) The heterogeneity analysis shows that the promotion effect of ESG responsibility fulfillment in TFP is more significant in the non-state-owned enterprises of the central region.

6.2. Revelations

Based on the conclusion that the fulfillment of ESG responsibilities will improve the TFP of textile enterprises, this paper gives the following initiatives: first, for the government, enterprises with good ESG performance should be given policy inclination, such as priority in loans and appropriate tax exemption; second, for enterprises, enterprises should vigorously promote green innovation and actively push forward the development of the production and R&D process of intellectualization and mechanization; third, enterprises should give full play to the creative ability of high-quality talents, and provide a platform for scientific and technological R&D to improve their subjective motivation; fourth, enterprises should actively disclose ESG information, which can reduce the existence of their research path with potential investors; fifth, enterprises should actively disclose ESG information. Finally, enterprises should give full play to the creative ability of high-quality talents, provide a platform for scientific and technological research and development, improve their subjective initiative, and support their scientific research. Enterprises should actively disclose ESG information, which can reduce the information asymmetry between them and potential investors, and also reduce the agency costs between owners and managers, thus enhancing the total factor productivity of enterprises.

6.3. Restrictions

Based mainly on Chinese-listed textile enterprises, this paper conducts empirical research on the relationship between ESG performance and total factor productivity, such as mediation tests, moderation tests, heterogeneity analysis, etc. Although the relevant data are processed as accurately as possible, and the arguments are sufficiently strong, this paper has the following shortcomings due to personal perception and other factors.

First, the research object is relatively singular, as the research sample only selected Chinese-listed textile enterprises; however, the textile industries in the United States, India, the United Kingdom, Vietnam, France, Italy, and other countries are more mature in development. Therefore, future scholars can further expand the research sample to provide more sufficient evidence for ESG performance in enhancing the total factor productivity of the textile industry.

Second, the mediation test and heterogeneity analysis are not perfect. This paper only studies the mechanism of green innovation and human capital level between ESG performance and total factor productivity in the textile industry; whether there are other paths is a direction that future scholars can explore. In addition, the heterogeneity analysis is only studied from the nature of the enterprise and geographic location; although it is in line with the direction of the majority of scholars’ research, it is expected that future scholars will be able to find more colorful and unique heterogeneity analysis indicators.