Economic Policy Uncertainty and ESG Performance: Evidence from China

. Using the data of listed companies in Chinese A-share market from 2011 to 2020, this paper investigates the effect of the economic policy uncertainty (EPU) on corporate environmental, social, and governance (ESG). The results show that during periods of high economic policy uncertainty, firms increase their overall ESG performance, corporate environmental performance, social performance and governance performance. Heterogeneous analyses show that the positive effect of EPU on ESG performance is more pronounced for state-owned enterprises, for firms with better corporate governance, for firms with more institutional investors, and for firms with less financing constraints. This study contributes to the literature on the determinants of ESG and provides implications for both practitioners and academics.


Introduction
With the focus on environmental issues and sustainability, corporate environment, social and governance (hereinafter referred to as ESG performance) is becoming increasingly important. Looming climate issues and COVID-19 have prompted a reconsideration of the sustainability factor. In addition to developed markets, emerging market economies have also increased their focus on corporate ESG in recent years. With the introduction of China's dual carbon strategy goal and the accelerated green transformation of the Chinese economy and society across the board, China's A-share market has been increasing its focus on ESG in recent years. However, the lack of a clear and explicit disclosure framework has led to the uneven quality of ESG information disclosure by Chinese companies. For emerging markets, improving their ESG disclosure is a key step to promote ESG development. Therefore, a study based on the Chinese Ashare market can first help developing markets improve ESG disclosure. Furthermore, studies have shown that economic policy uncertainty (hereinafter referred to as EPU) appears to be about one-third higher in developing countries than in developed countries (Bloom, 2016). Emerging markets are more vulnerable to exogenous uncertainty shocks than developed markets, and take significantly longer to recover (Carrière-Swallow and Céspedes, 2013). Given the growing importance of emerging markets in the globalized economy and their distinctive characteristics from developed countries, it is important to study how economic policy uncertainty affects economic activity in emerging markets. However, there are few studies on EPU and ESG based on developing markets. Vural-Yavas¸ et al. (2020) based on 15 developed European markets, tested and found a significant positive relationship between EPU and ESG. To my knowledge, this paper is the first study based on the China A-share market. As the world's largest developing market, a study based on the Chinese A-share market has two additional benefits. First, in a theoretical sense, it can fill the research gap between EPU and ESG in developing markets. Secondly, in a practical sense, it is not only beneficial for developing markets to deal with economic policy uncertainty, but also can contribute to further improving ESG disclosure in developing markets. This paper examines the relation between economic policy uncertainty and ESG performance in China during the period of 2011-2020 by using the Baker et al. (2013) index to measure China's EPU. Specifically, I find that EPU has a positive impact on corporate ESG performance. By decomposing its sub-dimensions, the results show that EPU promotes corporate environmental, social and corporate governance. From the perspective of property rights heterogeneity, the positive effects of economic policy uncertainty on overall ESG, environmental and governance are more salient for state-owned enterprises. In addition, internal governance attenuates the positive impact of EPU on overall ESG, and environmental and governance dimensions, while institutional investors strengthen the impact of economic policy uncertainty on ESG, environmental, social and governance. It is also found that high financing constraints attenuate the positive impact of EPU on ESG and social performance. This paper contributes to the existing literature in threefold. Firstly, this paper enriches the research on the consequences of economic policy uncertainty. Most of the existing studies reveal the darkness of economic policy uncertainty, while this paper confirms the positive relationship between EPU and ESG, which theoretically helps to view the event of economic policy uncertainty more objectively. Second, this study enriches the research on determinants of ESG performance. Most of the existing studies on ESG influencing factors focus on the micro level, but this paper enriches theoretically the research on the macro influencing factors of ESG. Finally, this paper investigates the impact of EPU on ESG and its sub-dimensions based on the Chinese A-share market, which makes a theoretical contribution to the study of the impact of economic policy uncertainty on ESG in emerging markets and fills the research gap. The remainder of this paper is structured as follows. Section 2 reviews related literature and develops our research hypotheses. Section 3 describes our data and methods. Section 4 reports results and findings. Section 5 concludes.

Literature Review The Consequences of EPU.
Most of the studies have found negative effects of economic policy uncertainty. Studies based on a micro perspective find that EPU increases the degree of information asymmetry between firms and external information users (Nagar et al., 2019), making it more difficult for firms to raise capital. Information asymmetry increases the risk faced by creditors and equity holders, which may lead them to demand higher returns or set stricter terms, thus increasing the cost of equity financing and debt financing for firms (Waisman et (Jensen, 2001;Borghesi et al., 2014). Some scholars also believe that ESG activities may provide opportunities for agency behavior of corporate insiders and negatively affect corporate governance (Masulis et al., 2015). However, with the popularization of the concept of sustainable development and the development of the concept of corporate governance, more scholars have begun to agree with the stakeholder theory, which states that maximizing corporate value should include stakeholder value in addition to shareholder value (Edmans, 2011). More studies show that ESG has a positive impact on company operations. First, CSR can be used as a form of insurance for companies to protect corporate value (Borghesi et al., 2019). Second, ESG activities can mitigate the risks to which companies are exposed (

Hypotheses Development
First, in times of economic policy uncertainty, companies can improve their ESG performance to help boost corporate performance. EPU can bring negative shocks to firm performance and reduce firm value. On the one hand, Borghesi et al. (2019) found that ESG can act as an insurance measure to accumulate social capital for firms and protect corporate value. On the other hand, ESG performance can help firms gain a reputational advantage in the market (Oussama et al., 2022), which in turn can improve customer loyalty and business continuity, positively impacting firm performance (Le et al., 2021; Aditya and Hananto, 2022;). Therefore, firms may enhance their operating performance and cope with the risks associated with economic policy uncertainty by improving their ESG performance. Second, EPU increases a firm's cost of equity capital and cost of debt ( . Therefore, companies with good ESG performance have a higher level of trust with their customers and can generate a more stable partnership. In order to maintain good relationships with stakeholders, companies will actively improve their ESG performance in times of high uncertainty. Combining the above analysis, derive the hypothesis H1. H1：Economic policy uncertainty has a positive impact on corporate ESG performance. In China, compared to Non-State Owned Enterprises (hereinafter referred to as Non-SOEs), State Owned Enterprises (hereinafter referred to as SOEs) have more economic, social and political responsibilities, and are subject to stricter regulations. As early as 2009, Chinese State-owned Assets Supervision and Administration Commission of the State Council required all SOEs to publish their CSR sustainability reports within three years, and in 2012, they established a Social Responsibility Steering Committee, requiring all SOEs to publicly release their CSR reports by June each year. First, EPU may exacerbate information asymmetries, but strict regulatory requirements for SOEs to disclose their ESG performance can facilitate SOEs to improve their ESG performance. In addition, SOE executives are political in nature, in times of economic policy uncertainty, SOEs actively fulfill their social responsibility, which helps executives' future personal development. Therefore, the impact of economic policy uncertainty on ESG may be greater in SOEs compared to non-SOEs. This brings us to our hypothesis 2. H2：Other things being equal, the positive impact of EPU on ESG is greater in SOEs compared to non-SOEs.
EPU motivates firms to be more active in ESG activities, which, in terms of agency theory, facilitates managers' agency behavior. In firms with poorer internal governance, managers are more likely to engage in self-interested behavior, which diminishes the positive impact of EPU on ESG. This leads to hypothesis 3. H3: Other things being equal, the positive impact of EPU on ESG is greater in firms with good corporate governance. Institutional investors play an external monitoring role for ESG. The study finds that institutional investors reap financial and social rewards by urging companies to improve their ESG profiles. Economic policy uncertainty increases the downside risk of stocks, and institutional investors have an incentive to urge firms to improve their ESG performance to avoid damage to their interests. Thus, in periods of high uncertainty, firms with more institutional investors are under more pressure from institutional investors to improve their ESG. Therefore, institutional investors may play a positive role between EPU and ESG. Hypothesis 4 is proposed. H4：Other things being equal, the positive impact of EPU on ESG is greater in firms with more institutional investors. Since ESG activities require significant capital investment by firms, the impact of EPU on ESG also varies depending on the degree of financing constraints faced by firms. When uncertainty is high, firms have an incentive to invest in ESG activities to enhance their ESG performance. Thus, when the degree of financing constraints faced by firms is low, firms are able to obtain more capital to perform ESG activities. Therefore, the degree of financing constraint may attenuate the impact of EPU on ESG. Derive hypothesis H5. H5：Other things being equal, the positive impact of EPU on ESG is greater in firms with less financing constraints.

Sample
This paper constructed a sample of Chinese A-listed firms, collecting data from 2011 to 2020. Our data on ESG were obtained from the Bloomberg database, and EPU is from Baker et al. (2013). All other financial data in this article are from the China Stock Market & Accounting Research (CSMAR) database. This paper exclude samples in financial industry, samples of ST (special treatment) firms, samples of PT (particular transfer) firms, and sample of firms at risk of delisting (*ST) and in the delisting process. The final sample consists of 1,110 unique firms for a total of 9,416 firm-year observations. To control for the effect of parameter extremes, this paper winsorized continuous variables at 1% and 99% levels.

Empirical Model
To examine the impact of economic policy uncertainty on the performance of ESG, building the model as follows: F.ESG i,t =β 0 +β 1 EPU i,t-1 +∑Controls i,t-∑Controls i,t-1 represents all control variables, the specific variables and their definition methods are explained in the following 3.3. ∑Year denotes control year fixed effects, meanwhile, ∑Industry represents the industry fixed effect, which is controlled according to the broad industry categories in the Guidelines for the Industry Classification of Listed Companies (2012 Revision) issued by the China Securities Regulatory Commission.
To alleviate endogenous problems, this paper lagged all the explanatory variables by one year.

Variables ESG
This paper use Bloomberg disclosure scores to measure corporate ESG performance. The Bloomberg ESG score was chosen for two reasons: First, the Bloomberg ESG score is more accurate. While many organizations base their ESG scores on subjective judgments and analyst rankings, Bloomberg ESG scores are based on a robust quantitative approach with data-based metrics. Second, Bloomberg ESG scores have a broader coverage. For example, only a few companies in the world currently disclose their greenhouse gas emissions, but Bloomberg leverages the advantages of its machine learning and natural language processing technology to estimate a company's greenhouse gas emissions based on existing disclosures and data, and incorporate them into the environmental metrics score a . Bloomberg takes into account the environmental, social and governance aspects of a company to obtain a composite score. This paper uses the natural logarithm of the Bloomberg score to measure ESG. in China that had to meet the following conditions: {{policy OR spending OR budget OR political OR "interest rates" OR reform} AND {government OR Beijing OR authorities}} OR tax OR regulation OR regulatory OR "central bank" OR "People's Bank of China" OR PBOC OR deficit OR WTO. They then apply these requirements in an automated search of every SCMP article published since 1995 to generate frequency statistics for articles on policy-related economic uncertainty. Finally, they divided the monthly frequency counts by the number of all SCMP articles in the same month, and normalized the index to 100 for January 1995. In this paper, EPU is calculated as follows: (1) calculate the annual mean of the monthly index of China's economic policy uncertainty, (2) to eliminate order-of-magnitude differences, then divide this annual mean by 100.

Control Variables
The control variables in this paper including company size (Size), financial leverage (Lev), company age (Age), company growth (MB), return on assets (ROA), ratio of independent directors (InRatio), board size (Board), enterprise value (TobinQ), shareholding concentration (Top1), and enterprise's proprietorship (SOE). Size is measured as a logarithmic measure of the firm's total assets. Lev is calculated as the ratio of a firm's liabilities to its total assets. Age represents the number of years the company has been listed. MB is measured by dividing total assets by market capitalization. ROA is calculated by dividing the company's current net income by its total assets at the end of the period. InRatio is the ratio of the number of independent directors to the number of board of directors. Board represents the natural logarithm of the number of board members. TobinQ is Tiobin's Q of the company. Top1 represents the percentage of shares held by the first largest shareholder. SOE of state-owned enterprises takes the value of 1, otherwise it is 0.

Descriptive Statistics and Correlation
Coefficient Panel A of Table 1 reports descriptive statistics for our main sample. First, the average ESG scores are 2.981, with a standard deviation of 0.313. The mean and median of G (corporate governance) scores are higher than both the E (environmental) and S (social) performance scores. The results show that the standard deviation of EPU is 2.371, which indicates large variation in EPU during the sample period. Panel B of Table 1 reports the Pearson's correlation coefficients between the main study variables, the maximum value of the absolute value of the correlation coefficient is 0.531, and the VIF values are all less than 10, indicating that there is no covariance problem.

Baseline Regression Results
This paper first examined the effect of EPU on ESG. Table 2 column (1) reports regression results that only control for year-and industry-fixed effects. The estimated coefficient of EPU is positive and statistically significant with 1% significance level, with a value of 0.029, which implies that one standard deviation increase in the EPU is associated with a 0.029 increase in ESG disclosure scores. Next, I reestimate the model after considering the control variables, the results are reported in Table 2 Column (2). The estimated coefficient of EPU is still positive and significant, with a value of 0.017. Also, to mitigate the potential estimation bias caused by firm-level heterogeneities, I reestimate our model by introducing firm-fixed effects, and the results are demonstrated in Table 2 Columns (3). The estimated coefficient also has a positive estimate of 0.031, with a t-statistic of 10.08. Further, to examine the effects of EPU on E (environmental), S (social), and G (governance), the results are demonstrated in Table 2 Columns (4) to Columns (6). The coefficients in columns (4), (5) and (6) are positive and significant, showing that the EPU promotes companies to improve their environmental, social and governance performance. All of these results are consistent with our previous estimation. The results suggest that firms enhance their ESG performance in response to economic policy uncertainty. The empirical results confirm hypothesis H1 and demonstrate that EPU not only promotes overall ESG performance, but also positively contributes to its environmental (E), social (S) and governance (G) sub-dimensions.

Further Analyses Heterogeneous Impact of Property Rights.
In order to investigate the potential impact of ownership structure, this paper introduce the interaction term between the SOE and EPU. The coefficient of the interaction term in Table 3 Columns (1) is significantly positive. The result demonstrates that the effect of economic policy uncertainty on ESG is more significant in state-owned enterprises. Hypothesis H2 is proved. To test the subdimensions of each ESG category, the coefficient of the interaction term in Table 3 Columns (2) and (4) are significantly positive, which shows that the effect of EPU on environmental (E) and governance (G) are more significant in state-owned enterprises. And Table 3 Columns (3) shows that the effect of EPU on social (S) is not significantly different between SOEs and non-SOEs. Note: Using one-period lagged independent variables to mitigate the endogenous problems. The numbers in parentheses are t-statistics. *p <.1, **p < .05, ***p < .01. Abbreviations: ESG, environmental, social, and governance practices; E, environmental; S, social; G, governance; EPU, economic policy uncertainty.

Heterogeneous Impact of Corporate Governance.
The level of corporate governance affects the impact of EPU on ESG performance. Using the management expenditure ratio (MER) to measure the corporate governance. The larger the MER, the worse the corporate governance. To construct the interaction term between MER and EPU to examine the effect of corporate governance between EPU and ESG performance. Table 4 Columns (1) shows that the coefficient of the interaction term is significantly negative. This result suggests that the positive impact of EPU on ESG is significantly weaker in firms with poorer corporate governance. The results are generally consistent with our conjecture laid out in hypothesis H3. Similar to the conclusion above, in companies with good corporate governance, the effect of EPU on environmental and governance are enhanced. Note: Using one-period lagged independent variables to mitigate the endogenous problems. The numbers in parentheses are t-statistics. *p <.1, **p < .05, ***p < .01. Abbreviations: ESG, environmental, social, and governance practices; E, environmental; S, social; G, governance; EPU, economic policy uncertainty. Heterogeneous Impact of Institutional Investors. This paper use the institutional investors ratio (IIR) to measure the firm's institutional investors. To construct the interaction term between IIR and EPU to examine the effect of institutional investors between EPU and ESG performance. Table 5 Columns (1) shows that the coefficient of the interaction term is significantly positive. This result suggests that the positive impact of EPU on ESG is significantly stronger in firms with more institutional investors. Hypothesis H4 is proved. Take it a step further, regressing E, S and G with the interaction term respectively, the results are shown in Table 5 Columns (2) to Columns (4). I find that the coefficients of the interaction term are still significantly positive in companies with more institutional investors. This means that in times of high uncertainty, institutional investors can not only contribute to the overall ESG performance of a company, but also positively contribute to environmental (E), social (S) and governance (G).  Table 6 Columns (1) shows that the coefficient of the interaction term is significantly negative. This result suggests that the positive impact of EPU on ESG is significantly stronger in firms with less financing constraints. Hypothesis H5 is proved. Take it a step further, regressing E, S and G with the interaction term respectively, the results are shown in Table 6 Columns (2) to Columns (4). The results show that high financing constraints reduce the positive impact of EPU on social, but not on environmental and governance. R-squared 0.261 0.188 0.136 0.236 Note: Using one-period lagged independent variables to mitigate the endogenous problems. The numbers in parentheses are t-statistics. *p <.1, **p < .05, ***p < .01.

Conclusion
This paper investigates the relationship between EPU and ESG performance in China over the period 2011-2020 by using Baker et al.'s (2013) index to measure economic policy uncertainty in China based on the Chinese A-share market. The study finds that EPU has a positive effect on corporate ESG performance. Examining its subdimensions, the results show that EPU promotes corporate environmental, social and corporate governance. From the perspective of property rights heterogeneity, the positive effects of economic policy uncertainty on overall ESG, environmental and governance are more salient for stateowned enterprises. In addition, internal governance attenuates the positive impact of EPU on overall ESG, and environmental and governance dimensions, while institutional investors strengthen the impact of economic policy uncertainty on ESG, environmental, social and governance. It is also found that high financing constraints attenuate the impact of EPU on ESG and social performance. The main findings of this paper suggesting that in emerging markets, economic policy uncertainty positively affects corporate ESG performance. This not only contributes to the development of measures to cope with economic policy uncertainty in developing countries, but also provides a theoretical basis for further improving ESG disclosure systems in developing countries. In addition, this paper demonstrates that institutional investors play a positive role between EPU and ESG, suggesting that effective supervision can promote corporate ESG performance in emerging markets, and that regulators should continuously improve the combination of government supervision and institutional supervision to promote healthy and sustainable market development. It is important to note that this study has two major limitations that can be addressed in future research. First, this paper focuses on a single country for the study, and the external ability may be relatively low. This paper finds a positive relationship between EPU and ESG based on the Chinese A-share market, and this finding can prove to some extent the positive impact of economic policy uncertainty on ESG in emerging markets, but the explanatory power is relatively weak. Future studies can be conducted based on multiple emerging markets. Second, this paper does not clearly distinguish between ESG and CSR-related literature when conducting the literature review. Due to the lack of literature on ESG, similar to the existing literature, CSR-related literature is partially cited in this paper when studying the impact factors of ESG and its economic consequences. With the enrichment of the literature, the boundary between CSR and ESG research should be further clarified in future studies.