Executive Incentives Matter for Corporate Social Responsibility under Earnings Pressure and Institutional Investors Supervision

: This paper theoretically explores the impact of the incentive preferences of executives (i.e., short-term incentives and long-term incentives) on corporate social responsibility (CSR) decisions (i.e., institutional CSR and technical CSR). Further, the paper presents the mechanism through which executives inﬂuence CSR activities by the pressures from ﬁnancial analysts and institutional investors supervision. Using a large sample of China-listed ﬁrms over 2007–2017, we achieve some helpful empirical results. The executives with short-term incentives tend to implement technical CSR strategy, while those with long-term incentives tend to implement institutional CSR strategy. Executives with short-term incentives, compared with those with long-term incentives, show stronger inter-temporal tradeo ﬀ s behaviors in the earnings pressure context. Furthermore, dedicated institutional investors can e ﬀ ectively attenuate the hypocritical behaviors of executives, and the e ﬀ ectiveness of governance shows a positive relationship with investors’ horizon. Our ﬁndings enrich the understanding on the relationship between the executives and CSR decisions in the earnings pressure context and further helps to perfect the institutional design in China’s listed


Introduction
On 19 August 2019, Business Roundtable announced a new statement signed by 181 CEOs who commit to lead their companies for the benefit of all stakeholders-customers, employees, suppliers, communities, and shareholders. (Since 1978, Business Roundtable has periodically issued Principles of Corporate Governance). The new statement focuses on creating long-term value for all stakeholders and prioritizing corporate social responsibility (CSR), which will result in shared prosperity and sustainability for all society. In China, as an emerging market, the Shenzhen Stock Exchange initially issued guidelines on the social responsibility of listed companies in 2006. Then, the relevant monitoring and evaluating clauses were presented to boost CSR reporting to balance the relationship between the extensive economic growth and environmental destruction [1]. More and more listed firms have taken CSR programs to demonstrate their good corporate citizenship [2]. However, many executives of China's listed firms are also severely lacking in their knowledge reserves related to the sustainable development strategy for society. Sometimes, CSR-related activities may be utilized to serve the interests of executives themselves. Hence, a more integrative framework needs to be proposed to examine CSR issues and to investigate these issues in the context of emerging economies.
The current literature focuses on the motivation of the CSR strategy implemented by executives from the two following competitive points [3,4]. First, for enhancing the harmonious relationship with the community and natural ecology, CSR activities can make the interests of executives, shareholders, and other non-financial stakeholders better aligned with firms [5]. Porter et al. [6] claim that "corporate success and social welfare are not a zero-sum game". Second, CSR activities are a form of wasteful spending, with the primary goal of enhancing the manager's private benefits at the expense of shareholders [7,8]. For example, insider-initiated corporate philanthropy is often engaged in by the top managers' own desires, not by the stakeholders' willingness.
However, the above literature neglects the analysis of why the executives carry out CSR activities or not and what motives executives' CSR decisions. Although some studies have explored the executives' CSR decisions from the views of education [9], age and tenure [10], and organization identification [11], these are only the personal features of executives. In fact, the executives' strategic decision, e.g., CSR decision, is a kind of interactive behavioral decision mostly influenced by external pressures. First, some studies show that the executives' decision can be affected by earnings pressure from financial analysts. The reason is that the investors believe in the prospect information of publicly listed firms from financial analysts, who are wise and knowledgeable experts [12]. Thus, executives of listed firms pay more attention to the financial analysts' earnings forecasts [13,14]. In order to attract the analysts' attention, executives usually perform inter-temporal behaviors to maintain high yield and share price of firms. This kind of purposeful intervention may limit investment in CSR-related activities, which is contradictory to the sustainability prospect of firms. Furthermore, the executives' decision also can be affected by the pressure from institutional investors' supervision. Since the executives cannot make decisions arbitrarily, they are constrained by the different institutional ownership structures [15]. It shows that the inter-temporal behaviors of executives may be restrained by the supervision of institutional investors [16]. These findings are conducive to understanding executives' actions on the strategic decision and inter-temporal tradeoffs. However, the existing researches pay less attention to the pressures of financial analysts and institutional investors on CSR decision by the executives. In 2019, Qian et al. [17] presented a helpful attempt to examine the impact of financial analysts on a firm's corporate social performance. They neglect the impact of the internal incentives of executives on CSR decisions. Motivation theory has proved that internal incentives, such as salary, bonus, and stock options, deeply affect executives' behavior. Hence, it is essential to construct an integrative framework, which can explain the relationship between the incentive preference of executives and CSR strategy when considering the pressures from financial analysts and institutional investors' supervision.
The main purpose of this paper is to provide an analyzing framework to conceptualize the impact of the incentive preferences of executives (i.e., short-term incentives and long-term incentives) on CSR decisions (i.e., institutional CSR and technical CSR) under the pressures from financial analysts and institutional investors supervision. We extend the work of Qian et al. [17] by introducing internal incentives and external pressure situations. We use a sample of China A-share-listed firms over 2007-2017 to study the issues discussed above, which would help to discover the status of China's corporate governance.
The contributions of this paper are presented as follows. First, this paper classifies the incentive preferences of executives as short-term incentives and long-term incentives. We investigate the relationship between these two different preferences of executives and the CSR strategy decision (e.g., institutional CSR and technical CSR). Our findings show that executives with long-term incentives prefer to choose institutional CSR items to enhance the harmonious relationship with the community, while those with short-term incentives tend to choose technical CSR items. Second, this paper attempts to incorporate the pressure from financial analysts into the theoretical framework explaining the relationship between the incentive preference of executives and CSR strategy decisions. Our findings can explain the reason why executives with short-incentives perform inter-temporal tradeoffs for short-term financial performance in the context of earnings pressure. Third, the pressures from institutional investors are divided into four distinct groups by portfolio concentration and investment horizon. They are integrated into the theoretical framework proposed above. The finding can present supervision effects from institutional investors when the executives with different incentive preferences make CSR decisions.
The remainder of this article is organized as follows. Section 2 reviews the relevant literature of this study and constructs a theoretical framework. Section 3 describes the data. Section 4 provides descriptive analysis and empirical results and briefly describes robustness tests to enhance the reliability of the conclusion. Finally, Section 5 concludes.

The Executive Incentives and Corporate Social Responsibility Strategy
According to the above literature, executives are essential participants of CSR decisions. In this paper, we investigate the preferences of executives, which can influence executives' CSR decisions. The executive incentives show a significant impact on the strategic decision of firms, including the CSR strategy. According to Finkelstein and Hambrick [18], executive incentives provide an effective way to measure the convergence of the interests of executives and shareholders. Specifically, compared with executives with short-incentives, these with long-incentives are more likely to converge with the interests of shareholders and pay more attention to capital investment [19], and are less sensitive to short-term financial performance [20]. Some evidence also supports that executive incentives could affect the CSR strategy. For example, Fabrizi et al. [21] find that both monetary and non-monetary incentives have an effect on CSR decisions. Hong and Minor [22] find that executives with direct incentives for CSR strategy are a useful tool to increase firm social performance.
According to the research of Souder and Bromiley [23], the firm s operating performance related to executive incentives can result in the agent problem because of incentive preferences mismatch. For example, the executives with short-incentives allocate more resources to projects and schemes that promote short-term performance and serve the interests of executives. These activities are not beneficial to the sustainable development of firms. Oh et al. [24] point out that older CEOs may be less likely to perform long-term investments, e.g., CSR strategy, and more likely take short-term profit-generating initiatives. Antia et al. [25] also show that short-term executive incentives are an essential indicator of investment to chase faster returns. Marinovic and Varas [20] note that executives who are more sensitive to short-term performance are more likely to engage in performance manipulation. Therefore, we need to classify executive incentives when implementing CSR strategy. The executives with short-term incentives incline to curtail the use of CSR strategy to promote financial performance in the short term due to their career security and reputation. In contrast, executives with long-term incentives pay more attention to the sustainable development of the firm and are more likely to implement CSR strategy because CSR-related activities could promote the long-term value and sustainable development of firms in the long run.
Based on the discussion above, this paper proposes the following hypothesis: Hypothesis 1a. The short-term incentives of executives show a negative relationship with CSR strategy.
Hypothesis 1b. The long-term incentives of executives show a positive relationship with CSR strategy.

The Executive Incentives and CSR Strategy when Considering Earnings Pressure from Financial Analysts
Earnings pressure may result in an earnings expectation gap between financial analysts and executives [26]. Some of the literature supports the idea that earnings pressure stimulates myopic behavior in executives, such as earnings management [24], and increases the current earnings at the expense of future earnings [17]. These executives attempt to weaken earnings pressure by cutting strategic investments [27,28] and corporate downsizing [29]. Duong and Pescetto [30] also show that earning management driven by earnings pressure may be implemented to maintain high duration valuations of the market.
In addition, some studies show that the reaction of executives to earnings pressure depends on their different characteristics and equity institutions. For example, compared with executives with short-term incentives, these with long-term incentives are less responsive to earnings pressure [15]. Lee and Chang [31] point out that executive incentives are likely to affect the reactions of executives to earnings pressure. Moreover, ownership structure shows a significant impact on the response of executives. Schulz and Wiersema [29] show that institutional investor stock ownership weakens executives' sensitivity under market pressure. Generally, the earnings pressure plays an essential role in investigating the impact of executive incentives on CSR strategy.
Based on the discussion above, this paper puts the following hypothesis: Hypothesis 2a. The short-term incentives of executives show a more negative relationship with CSR strategy under earnings pressure.
Hypothesis 2b. The long-term incentives of executives show an insignificant relationship with CSR strategy under earnings pressure.

The Executive Incentives and CSR Strategy when Considering Pressure from Institutional Investors Supervision
As an effective monitoring power in emerging markets, institutional investors play an important role in restraining executive short-termism and myopic behaviors [32]. According to the research of Gillan and Starks [33] and Connelly et al. [16], they point out that interest convergence between executives and shareholders can be gained through multiple governance mechanisms. For example, Alvarez et al. [34] find that institutional investors can influence the investment decisions of executives. The institutional shareholders can promote the implementation of long-term strategy by executives and CSR practices of enterprises, improving investor wealth [35]. Moreover, Fich et al. [36] show that the monitoring attention of institutional investors depends on portfolio positions. Dedicated investors have more incentives to monitor the behaviors of executives and show greater tolerance for short-term performance disappointment [37]. Inversely, transient investors frequently trade in and out of firms on the base of change in stock market value and have fewer incentives to behaviors of executives [38]. Li et al. [39] provide evidence that institutional shareholders with short-term horizon are more likely to loosen regulatory constraints. Inversely, institutional shareholders with long-term horizon show an essential supervision function in restraining myopic managerial behaviors [40]. However, there are few studies that explain the relationship between institutional shareholders with short-term horizon when considering the governance function of institutional investors. Therefore, this paper provides interesting governance hypotheses considering the portfolio and investment horizon of institutional investors, which can help to clarify the governance mode of distinct classification groups.
Based on the discussion above, this paper proposes the following hypothesis: To investigate the complex relationship between the incentive preference of executives and CSR strategy under the scenario of the earnings pressure and institutional investors, this paper constructs the theoretical framework, as shown in Figure 1. In this theoretical framework, the research process can be divided into three stages.

Theoretical Framework
To investigate the complex relationship between the incentive preference of executives and CSR strategy under the scenario of the earnings pressure and institutional investors, this paper constructs the theoretical framework, as shown in Figure 1. In this theoretical framework, the research process can be divided into three stages.

Corporate Social Responsibility
The measurements of CSR performance in this study are based on the Chinese Research Data Services Platform (abbr. CNRDS), providing the CCSR database. (The databases providing the measurement of CSR performance of China listed firms are the CNRDS, RKS, and HEXUN Web CSR rating dada in China. This paper takes CNRDS as the primary database to measure the CSR scores because the CSR scores from CNRDS and RKS show significant consistency, and CNRDS could provide more qualitative indicators than RKS.) The CCSR database takes the KLD STATA database as the mainframe model and provides various indicators of CSR measurement based on the disclosed annual CSR reports by the Shanghai Stock Exchange (SSE) and Shenzhen Stock Exchange (SZSE) since 2006. CCSR evaluates CSR performance of Chinese listed firms based on six qualitative indicators (CCSR provides the six qualitative indicators about CSR performance, which are the community, corporate governance, diversity, employee relations, environment, and product quality and safety, respectively) and statistics annually from two dimensions of strengths and concerns.
This paper divides the CSR-related activities into institutional CSR and technical CSR (i.e., ICSR and TCSR) according to the method of Freeman et al. [41]. The former refers to some responsible activities serving the outside stakeholders of enterprises, such as community relations, environmental stewardship, and diversity. (Institutional CSR is more likely to generate moral capital, reduce the enterprise's financing constraints, and effectively buffer or minimize the damage to the corporate image from negative exogenous events in theory [10].) The latter refers to some responsible activities serving the inner stakeholders of enterprises, such as employee relations, product safety, and governance. (The technical CSR is more likely to generate less moral capital, which may embed the interests of the executives into the inner CSR activities [42,43].) Following Mattingly and Berman [9] and Godfrey et al. [44], this paper defines INCSRS and INCSRNC as the sum scores of the strengths and concerns items under the institutional CSR category, and TECSRS and TECSRC as the sum scores of the strengths and concerns items under the technical CSR category, respectively.

Corporate Social Responsibility
The measurements of CSR performance in this study are based on the Chinese Research Data Services Platform (abbr. CNRDS), providing the CCSR database. (The databases providing the measurement of CSR performance of China listed firms are the CNRDS, RKS, and HEXUN Web CSR rating dada in China. This paper takes CNRDS as the primary database to measure the CSR scores because the CSR scores from CNRDS and RKS show significant consistency, and CNRDS could provide more qualitative indicators than RKS). The CCSR database takes the KLD STATA database as the mainframe model and provides various indicators of CSR measurement based on the disclosed annual CSR reports by the Shanghai Stock Exchange (SSE) and Shenzhen Stock Exchange (SZSE) since 2006. CCSR evaluates CSR performance of Chinese listed firms based on six qualitative indicators (CCSR provides the six qualitative indicators about CSR performance, which are the community, corporate governance, diversity, employee relations, environment, and product quality and safety, respectively) and statistics annually from two dimensions of strengths and concerns.
This paper divides the CSR-related activities into institutional CSR and technical CSR (i.e., ICSR and TCSR) according to the method of Freeman et al. [41]. The former refers to some responsible activities serving the outside stakeholders of enterprises, such as community relations, environmental stewardship, and diversity. (Institutional CSR is more likely to generate moral capital, reduce the enterprise's financing constraints, and effectively buffer or minimize the damage to the corporate image from negative exogenous events in theory [10]). The latter refers to some responsible activities serving the inner stakeholders of enterprises, such as employee relations, product safety, and governance. (The technical CSR is more likely to generate less moral capital, which may embed the interests of the executives into the inner CSR activities [42,43]). Following Mattingly and Berman [9] and Godfrey et al. [44], this paper defines INCSRS and INCSRNC as the sum scores of the strengths and concerns items under the institutional CSR category, and TECSRS and TECSRC as the sum scores of the strengths and concerns items under the technical CSR category, respectively. Besides, this paper calculates the net CCSR scores to measure ICSR and TCSR level, denoted by NEINCSR and NETECSR, respectively.

Executive Incentives
Based on the studies of Lee et al., [42] and Cao et al., [45], this paper classifies the executive incentives as the short-term incentives and the long-term incentives. Short-term incentives for executives (SHOEXE) are measured by the natural logarithm of the sum of salary, bonus, and other cash payments of the top three senior executives. When the financial performance of enterprises is better at the end of the year, the executives will obtain more in terms of salary, bonus, and other cash payments. Therefore, executives with relatively short-term incentives are more sensitive to short-term financial performance and thus pay more attention to short-term earnings rather than long-term performance.
Moreover, long-term incentives for executives (LONEXE) are measured by the natural logarithm of the market value of the shares and the unexercised stock options of the top three senior executives. Therefore, executives with relatively long-term incentives pay more attention to impression management and are more likely to implement CSR strategy and thus are inclined to promote future financial performance.

Earnings Pressure
This paper adopts the method of Zhang and Gimeno [26] to calculate the proxy of earnings pressure (EARPRE). Contrasted with other measures of earnings pressure, this method provides the reliability where this variable could not be affected by the endogenous events, such as earning management. Earnings pressure could be measured for firm i at the beginning of year t as where Analyst consensus forecast denotes the mean of the earnings forecasts from financial analysts at the beginning of reporter term [46] and Potential earnings denotes the potential performance of firms at the beginning of reporter term [47]. Zhang and Gimeno [26] point out that the potential performance of executives could be influenced by the historical performance of firms and industry performance over the same period. Thus, potential earnings are calculated by historical performance containing information of stock price before the analysts' forecasts, and the average peer potential earnings performance. (When calculating the variable of potential performance, this paper excludes the focal firm, and controls the same earnings benchmark at the same industry for the accounting year). This paper sets the weights of each part above to 0.6 and 0.4, respectively. (We also set the weight of historical earnings and peer earnings to 0.5, respectively, and the results are robustness). Table 1 shows, the mean differences between analyst forecast consensus and reported earnings are significantly higher than that between potential earnings and reported earnings (paired t-test: p < 0.01). In comparison, the estimation of potential earnings is unbiased (p > 0.1). The distribution of analyst forecast consensus at the beginning of the reporter term is skewed toward overestimating reported earnings (with a mean difference significantly above 0). The results shown in Table 1 are consistent with previous literature. Thus, analyst forecasts are valuable in our study, even if they appear to be biased.

Institutional Investors
Following the approach by Fich et al. [36], this paper captures the relative importance of the target firm concerned by the institutional investors. To estimate the market value owned by the institutional investors, the calculation is that the firm's market values multiply the percentage of the firms' equity held by the institutional investors. Then, we calculate the rate of the market value held by the institutional investor in a single firm to the value of institutional investor's portfolios. At last, we define dedicated institutions (DEDIINS) as those whose holding value in the target firm is in the top 10% of their portfolios, and transient institutions (TRANINS), otherwise.
In addition, this paper also calculates the proxy of the investment horizon. Since the proxy of the investment horizon cannot be directly observed, it is measured by constructing the indicator about the frequency of investor trades. According to the research of Gaspar et al. [48], the main proxy for the investment horizon is the average holding time, which is equal to the value of the internal term divided by investment turnover where the internal term is 1. (See Appendix B) The lower investment turnover thus shows that the firm is held by investors with long-term horizon (LONINS). Similarly, the higher investment turnover thus shows that the firm is held by investors with short-term horizon (SHOINS).
This paper constructs four institutional investor variables depending on the median values of the portfolio concentration and investment horizon. They are dedicated investors with long-term horizon (LONDEDI), dedicated investors with short-term horizon (SHODEDI), transient investors with long-term horizon (LONTRAN), and transient investors with short-term horizon (SHOTRAN).  [26], we design the experiment of the validation of the earnings pressure measure in our study). I Both deviation 1 and 2 were the focal estimate minus reported earnings. N = 10532 (firm-levels). II The first two tests are one-sample t-tests; the third is a two-sample t-test. *** p< 0.01 One-sided tests. Table 2 reports the summary statistics about the variables used in the paper (in addition to the CCSR, CMSAR, and WIND database, several other databases are used in this study. We also use the CCSR economic database and the RESET financial database to fill in the missing values of the firm-level variables), including CSR scores, executive incentives, earnings pressure, and other several firm characteristics. All non-dummy variables are winsorized at the first and 99th percentiles, and Appendix A explains their constructions.

Descriptive Statistics
In particular, the average NETINCSR and NETECSR scores are 6.147 and 10.841. After distinguishing these scores into strengths and concerns item scores, the average scores of INCSRS and TECSRS are 3.305 and 10.893, while the average scores of INCSRC and TECSRC are 0.239 and 0.127, respectively. (Considering the measure differences in dimensions of CSR performance, the CSR-related measures were placed in standardized processing to gain a dimensionless performance variable). Furthermore, the average scores of LONEXE and SHOEXE are 4.827 and14.007, indicating the motivation level of the long-term incentives and the short-term incentives of executives. Together, the average share is 4.066% (6.788%) in institutional ownership by DEDIINS (TRANINS) in our sample, which includes 3.677% of the total shares held by LONDEDI (3.086% held by LONDEDI) and 2.119% held by SHODEDI (1.785% held by SHODEDI). Similarly, the average shares in institutional ownership by LONINS (SHOINS) are 3.928% and 1.879%.

Main Results on Executive Incentives and CSR Types
In order to test H1a and H1b, we construct the following model: where XCSR_Score denotes the CSR scores of firms, and mana_incentives denotes the incentives of executives. X includes other control variables for several firm characteristics. Meanwhile, year fixed effects (ϕ) and industry fixed effects (η, CSRC industry code) are controlled in the model. What is more, the time interval between CSR strategy policy and executive incentives, independent variables with a lag of one period, are introduced into the model.   (2) and (5)-(6) in Table 3 show that executives with short-term incentives (SHOEXE) and executives with long-term incentives (LONEXE) are significantly positively associated with the CSR strengths items. Besides, the coefficients of CSR scores on the executive with short-term incentives are higher than the coefficients of executives with long-term incentives. Furthermore, the coefficients of TECSRS on executives with short-term incentives are higher than the coefficients of INCSRS on executives with short-term incentives. In contrast, the coefficients are reversed for executives with long-term incentives. For example, in column (1), the coefficient on SHOEXE is 1.626, which is less than the coefficient on SHOEXE in column (5) and is significant at the 1% level, with a t-statistic of 13.36.
Columns (3)-(4) and (7)- (8) in Table 3 show that executives with short-term incentives and executives with long-term incentives are significantly positively associated with the CSR concerns items. Both executives with short-term incentives and executives with long-term incentives are significantly positively associated with CSR strengths and concerns items, but the coefficient on CSR strengths is significantly higher than that on CSR concerns. The results show that executives are more likely to enhance the sustainable development of the enterprises, rather than hinder through CSR-related activities despite the CSR types despite depending on the executive incentives, which are consistent with Liu [49]. Our analysis reveals the fact that executives with short-term incentives are more inclined to choose technical CSR. The technical CSR could improve the company's internal governance, increasing firms' profit-making interest. This finding is opposite to the research of Souder and Bromiley [23]. This finding provides a reasonable explanation for the mixed results on the relationship between the executives and CSR decisions in previous literature. However, executives with long-term incentives prefer to choose institutional CSR to enhance the community relationship, which is consistent with Sanders [50].
Interestingly, this study shows that executives with short-term incentives preferred to implement CSR strategies than those with long-term incentives. The results are seemingly paradoxical to the literature regarding myopic executive, stating that short-incentives executives prefer to boost short-term earnings by reducing capital investment [51]. Particularly, this study empirically shows that short-term incentives executives have a strong preference for CSR strategies, which promotes the sustainable development of enterprises, not only for technical CSR but also for institutional CSR. The results demonstrate that the visibly ethical behaviors of insider-initiated CSR-related activities in China could be regarded as a particular case of manager building empires [52], an operation distress insurance [53] and corporate hypocrisy with self-interest embedded in benevolence [54].

Additional Hypocrisy Evidence on CSR Types
In this section, we further test the hypothesis that executives with short-term incentives may exacerbate hypocrisy and self-interest myopic, compared with executives with long-term incentives executives. We isolate the effect of different incentives for executives on CSR types by distinguishing earnings management, taking the absolute level of discretionary accruals as the sorting criterion. This study uses a modified version of the Jones model [55,56] to measure the proxy of earnings. (See Appendix D). Table 4 presents the results of executives with different incentives on two types of CSR by distinguishing earnings management. In columns (1) to (8) of Table 4A, the coefficients reveal that executives with short-term incentives could have a more significant positive effect on CSR strengths items, including INCSRS and TECSRS under earnings management. For example, the coefficient is 1.342 in column (1) with a t-statistic of 4.35 and 1% statistical significance, which is less than the coefficient in column (2). On the contrary, executives with short-term incentives could have a less significant positive effect on CSR strengths items, including INCSRC and TECSRC under earnings management. For example, the coefficient is 0.239 in column (3) with a t-statistic of 4.35 and 1% statistical significance, which is more than the coefficient in column (2). The coefficients show that executives with short-term incentives could disguise earnings management through CSR-related activities and embed self-interest into charity behaviors. The conclusions provide the evidence that CSR may be a hypocritical tool for executives to achieve more private gains at the expense of stakeholders, which is consistent with Arli et al. [57].
Similarly, Columns (1) to (8) of Table 4B show the consequences of executives with long-term incentives on the CSR types. Although executives with long-term incentives may conduct earnings management through CSR-related activities, their behavioral intensity is much less than that of executives with long-term incentives. This result provides evidence that CSR-related activities may be a hypocritical tool to be used to practice earnings management by executives for their reputation. This finding is not consistent with the research of Philippe and Koehler [58], who support that executives may blindly pursue financial performance and ignore ethical issues when pursuing short-term performance. Since in China, the life cycles of listed companies are shorter than European companies. In this regard, CSR-related activities are more likely to be used as a tool for serving the interest of executives and the short-term performance of the firm rather than the sustainable development of the firm.

Main Results on Executives with Different Incentives and Inter-Temporal Tradeoffs
This section investigates whether executives with disparate incentives could practice the inter-temporal tradeoffs for CSR-related activities under earnings pressure. In Table 5, the coefficients show that executives with short-term incentives have intensely motivated to practice inter-temporal tradeoffs under earnings pressure. This result is consistent with Qian et al. [17], who point out that earnings pressure stimulates the myopic behavior of executives and thus increases current earnings at the expense of future earnings. The inter-temporal tradeoffs are mainly reflected in technical CSR items, while the institutional CSR items are not distinct. For example, the interaction terms between executives with short-term incentives (SHOEXE) and earnings pressure (EARPRE) are significantly negative with technical CSR strengths items. The coefficient of interaction terms is −1.597 with a t-statistics of −2.07 and 5% statistical significance in column (5), while the coefficient of interaction terms on institutional CSR strengths items is not significant in column (1). The possible explanation may be that institutional CSR-related activities could enhance the accumulation of moral capital, and attenuate the negative effect harmful to the enterprises. Furthermore, the coefficients also show that executives with short-term incentives tend to take institutional CSR items as "insurance" for negative evens or earnings obsession. In particular, executives with short-term incentives may reduce CSR concerns activities under earnings pressure; thus, they could gain more "community reputation" or moral capital at less expense. The results provide the evidence that goodwill-related activities may attenuate the reaction of stakeholders to poor performance and adverse events, consistent with Lins and Servaes [59] and Flammer [60]. For example, the interaction terms between executive short-incentives (SHOEXE) and earnings pressure (EARPRE) are significantly negative with technical CSR strengths items. The coefficient of interaction terms is −0.158 with a t-statistics of −1.83 and 10% statistical significance in column (6), and the coefficient is −0.136 with a t-statistics of −1.69 and 10% statistical significance in column (2).
What is more, it is worth noting that executives with long-term incentives would like to make inter-temporal behaviors of weakening CSR strengths items under earnings pressure. Compared with other long-term strategies, reducing institutional CSR strengths could significantly promote the increase of short-term performance rather than introducing a negative implication. Although executives' long-term incentives focus on the sustainable development of enterprises by practicing long-term strategies, they still pay attention to the fluctuations of the stock price. This finding is consistent with Antia et al. [25]. Table 6 presents the results for hypocrisy and the inter-temporal tradeoffs for different CSR activities under earnings pressure. Columns (1) to (4) of Table 6 show the results of the inter-temporal tradeoffs of executives with short-term incentives in the context of low earnings management ("EM low"). The interaction terms between executive short-incentives (SHOEXE) and earnings pressure (EARPRE) show positive relationships with the net CSR scores, including NETINCSR and NETECSR, while statistical significance is more than 10%. Besides, columns (5) to (6) of Table 6 show the results in the context of high earnings management ("EM high"). The interaction terms between executive short-incentives (SHOEXE) and earnings pressure (EARPRE) show positive relationships with the net CSR scores, including NETINCSR and NETECSR. The coefficient of the interaction term is −3.731 with a t-statistics of −1.65 and 10% statistical significance in column (6), and the coefficient is −4.741 with a t-statistics of −1.99 and 10% statistical significance in column (8).  What's more, the coefficients of the interaction terms between executives with short-term incentives (SHOEXE) and earnings pressure (EARPRE) on technical CSR items (NETECSR) are higher than the coefficients of the interaction term on institutional CSR items (NEINCSR). The coefficients indicate that executives with short-term incentives are inclined to take technical CSR-related activities as "hypocritical tools" compared with institutional CSR -related activities. The results supplement the empirical evidence that executives take CSR-related actions as a tool to disguise hypocrisy with self-interest embedded in benevolence.

Do Institutional Investors Supervise the Inter-Temporal Tradeoffs for CSR Types of Executives under
Earnings Pressure?

Main Results on Institutional Investors and CSR Types of Executives
In this section, we would like to answer three questions. The first is whether institutional investors monitor executives to different CSR activities in the capital market. The second is how institutional investors with disparate characters affect the executives' choices on various CSR activities. The third is whether there is a significant change for executives' decisions on different CSR activities when considering the monitoring of institutional investors. Table 7 presents the results as follows. The coefficients reveal that the transient institutional investors (TRANINS) with stocks holding in numerous enterprises and frequently trading in and out of enterprises show opposition to CSR-related activities. The interaction terms between short-incentives (SHOEXE) and transient institutional investors (TRANINS) are significantly negative with institutional CSR items (NETINCSR) and technical CSR items (NETECSR). This result could be supported by the research of Fich et al. [36], who point out that transient investors frequently trade in and out of firms on the base of change in stock market value, and they are less likely to perform corporate governance. However, the absolute values of coefficients in the context of the S-TIH subsample are higher than the L-TIH subsample. For example, the coefficient of the interaction term is −0.128 with a t-statistics of −4.14 and 1% statistical significance in column (1), which is more than the coefficient in column (5). Similarly, the coefficient of the interaction term is −0.131 with a t-statistics of −3.97 and 1% statistical significance in column (2), which is higher than the coefficient in column (6).
In addition, the dedicated institutional investors (DEDIINS) with stocks holding in a small number of enterprises and non-frequently trading have incentives to supervise hypocritical behaviors, which is consistent with Cho et al. [40]. The coefficients of interaction terms between executives with short-term incentives (SHOEXE) and dedicated institutional investors (DEDIINS) is not significant, compared with the coefficient in column (4). For example, the coefficient of interaction terms is −0.205 with a t-statistics of −0.73 and more than 10% significance in column (3), and the coefficient is 0.468 with a t-statistics of 1.71 and 10% statistical significance in column (4). Although the coefficients of interaction terms are not significant, the dedicated institutional investors (DEDIINS) effectively balance the opposition to executives with short-term incentives' choice on CSR-related activities. Especially, the coefficient in column (4) shows that the dedicated institutional investors (DEDIINS) could promote the technical CSR items of enterprises. Table 8 presents the results for institutional investors and the inter-temporal tradeoffs of executives under earnings pressure in the context of investment horizon subsamples. The results provide additional evidence that institutional investors with the long-term horizon (LONINS) can effectively constrain executive hypocrisy. In particular, the investment horizon of institutional investors shows an inverse relationship with the level of the executives' support for CSR-related activities. To dig deeper into this effect, this section constructs the interaction term, including the executive incentives and earnings pressure in the context of investment horizon subsamples. This method helps to investigate the change for CSR preferences of executives when considering the monitoring of institutional investors. Notes: All models include year and industry fixed effects. *, **, and *** denote statistical significance at the 10%, 5%, and 1% levels, respectively.   Table 8 show that the interaction terms between executives with short-term incentives (SHOEXE) and earnings pressure (EARPRE) are significantly negative with the net institutional CSR items (NEINCSR) and the net technical CSR items (NETECSR), respectively. For example, the coefficient of the interaction on the net institutional CSR scores (NEINCSR) is −0.304 with a t-statistics of −1.90 and 10% statistical significance in column (2). The coefficient of the interaction on the net technical CSR scores (NETECSR) is −0.298 with a t-statistics of −1.74 and 10% statistical significance in column (4). The coefficients indicate that executives with short-term incentives practice inter-temporal tradeoffs under earnings pressure if institutional investors have a short investment horizon.

Institutional Investors and the Inter-Temporal Tradeoffs for CSR Types
Columns (5) to (8) of Table 8 show that the interaction terms between executives with short-term incentives (SHOEXE) and earnings pressure (EARPRE) show an adverse relationship with the net institutional CSR items (NEINCSR) and the net technical CSR items (NETECSR), respectively. Although the coefficient is not significant at the statistical level, the institutional investors with long-term horizon (LONINS) effectively balance the inter-temporal tradeoffs of executives. The result also provides additional evidence that the dedicated institutional investors (DEDIINS) with long-horizon could play important monitoring roles in correcting the CSR strategy of executives.

Endogeneity
The endogeneity concern could be the main problem in the above-related research. To rule out reverse causality, we employ the instrumental variables method of system moment estimation to avoid the endogeneity concern, and the results are presented in Appendix D. The most challenging task is the selection of instrumental variables, which require the strict method definition. In this paper, we employ the lagged stock concentration and the lagged investment concentration of institutional investors as the instrumental variables. The instrumental variables are likely to affect the size and duration of institutional investors' holdings, but impossibly directly affect the institutional and technical CSR scores. In the section of Table A2 in Appendix D, the study regresses the net CSR scores, including NETINCSR and NETTECSR, on institutional investors according to their portfolio durations and average holding time based on the model of two-step Generalized Method of Moments (GMM). The coefficient reveals that both transient and dedicated institutional investors tend to choose behavioral models that not supporting CSR activities. Especially, the strength of the opposition coefficients is significantly negative with the average holding time based on the perspective of the investment horizon. Above all, the results provide additional evidence that institutional investors have a causal effect on CSR, and the results in this study are robustness.

Alternative Measure for Executive Incentives
Although the selection and measurement of variables used in this study are discussed in detail, this study rechecks the robustness of our results employing the alternative measure for executive incentives. This research employs the absolute values of the proxies to describe executive incentives. Namely, when the executives are honored with compensation incentives or equity incentives by shareholders, the occurring of principal-agent problems would be decreased. Furthermore, in order to make sure our conclusion robustness, the executive incentives of the relative index are constructed. To classify executives according to the relative index, this study sorts the executive incentives into the relative long-term incentives and the relative short-term incentives by taking the median as the grouping criterion.

Further Robustness Tests
To ensure the robustness of the conclusions, this paper practices three basic checks. First, each regression with different samples by splitting the time, such as the period from 2010 to 2017, and the period from 2013 to 2017 is running. Second, the industry fixed effects based on different industry classification codes are recreated, including the companies less than two years. Third, the models, including firm fixed effects and province fixed effects, are re-estimated. The conclusions of this paper are robust to all three tests.

Conclusions
This study is a conductive supplement for the strand of literature on CSR activities in China as an emerging market. We conceptualize the impact of the executive incentives on CSR activities, then extend the framework to the case that firms are under pressure from the financial analyst and institutional investors. A series of robustness tests provide robust evidence for our conclusions. This paper has the following findings.
First, executives with different incentive preferences choose different CSR-related activities to enhance the sustainable development of the enterprises, rather than hinder firm growth due to their self-interests and myopic behaviors. Especially, executives with long-term incentives prefer to choose institutional CSR items to enhance the harmonious relationship with the community, while those with short-term incentives tend to choose technical CSR items. Moreover, executives with short-term incentives may disguise hypocrisy through practicing CSR-related activities and embed self-interest into charity behaviors. Furthermore, although executives with long-term incentives also perform earnings management for current stock price stability, they would like to practice CSR-related activities, especially institutional CSR.
Second, executives could perform inter-temporal tradeoffs for short-term financial performance in the earnings pressure context. This article suggests that executives with short-term incentives are more likely to practice inter-temporal tradeoffs under earnings pressure. The inter-temporal tradeoffs mainly appear in technical CSR items, not obviously for institutional CSR items. Furthermore, although executives with long-term incentives practice the inter-temporal tradeoffs, the intensity of their motivation is significantly weaker than that of executives with short-term incentives. It shows that executives with long-term incentives could pay more attention to the long-term value growth of enterprises while maintaining the short-term stock price stability.
Third, institutional investors play an important role in supervising the hypocrisy of executives and help to improve the inner governance of firms. Moreover, institutional investors with disparate features show different governance effects. Dedicated institutional investors help to restrain the hypocrisy behaviors of executives, who hold stocks in a small number of companies and non-frequently trade. Executives are less likely to practice inter-temporal tradeoffs in the earnings pressure context if institutional investors have a long-term investment horizon. In conclusion, dedicated institutional investors with long-term horizon show better governance effect in supervising and correcting behaviors of executives.

Managerial Implications
Our findings provide managerial implications for executives and investors. This article suggests that, in China, listed company investors should focus on the incentive preferences of executives to responsible investments, because the CSR-related activities may be a kind of corporate hypocrisy with self-interest embedded in benevolence. For example, executives with short-term incentives are inclined to take technical CSR-related activities as "hypocritical tools" rather than institutional CSR-related activities. These hypocritical behaviors could be harmful to the sustainable development of firms and be utilized to achieve executives' self-interests at the expense of shareholders' interests. Moreover, investors should pay attention to the earnings pressure on executives, because executives may practice inter-temporal tradeoffs to promote short-term performance by cutting long-term investment. These inter-temporal tradeoffs behaviors are not conducive to the sustainable development of firms due to their strategic uncertainty. Also, institutional investors play an important role in supervising and regulating the hypocrisy behaviors of executives. As a recent emerging supervision power in China, the types of investors may significantly affect the executive strategic decision and financial performance of the firm due to their strengths in capital and information. Therefore, it is helpful to improve the ownership structure of listed firms to enhance the governance effects of institutional investors, such as increasing shares of dedicated investors with long-term horizon.
Additionally, the results of this study can stimulate the interest of policymakers since it provides some evidence about the paradigms of executives and the incentive preferences of executives to CSR in the pressure situations. Our findings provide policy implications for policymakers in China as follows. First, some corporate governance regulations should be issued to perfect the institutional design of listed firms and to regulate the behaviors of executives with disparate incentives. Second, it's urgent to reduce information asymmetry between the firms and the investors to further promote transparency both in financial and non-financial information of listed firms in the emerging market. Third, because of the lack of reasonable policy guidance for institutional investors in emerging markets, it is beneficial to design reasonable multi-level guidance for different institutional investors.

Limitations and Future Research
This study presents the tests of the impact of the executives with different incentive preferences on CSR activities, then extend the framework in the pressure situations. Although the results of this article are valuable and provide some meaningful implications, several aspects limit the current research. First, the executive incentives as the proxy are utilized to measure the behavior paradigms of top-level executives, and it can't adequately represent the motivation of executives' behaviors due to the unreported compensations and hidden benefits existing. Therefore, the future search should construct a more effective proxy to executive incentives, which consider the unreported compensations and covered benefits. Second, the earning pressure is measured by the gap between the consensus of financial analysts and the potential performance of firms evaluated at the beginning of the reporter term. This measure ignores the changes in the market environment over the accounting year, which may affect the measure of this proxy. Future research may introduce the dynamic environment into the analysis framework. Finally, more dimensions for CSR strategies should be considered, which is constructive to enrich the results and implications of research in the emerging markets.
Author Contributions: The authors contributed equally to this work. All the authors contributed to the conceptualization, formal analysis, investigation, methodology, writing of the original draft, and review and editing. All authors have read and agreed to the published version of the manuscript.
Funding: This research was funded by the National Social Science Fund of China (19VHQ002).

Conflicts of Interest:
The authors declare no conflict of interest.  The natural logarithm of the total compensation of the top three senior executives plus one. Source: WIND database.

SHOEXE Short Executive Incentives
The average market value of restricted stock and unexercised stock options, owned by the senior executives. Source: WIND database.

DEDIINS Dedicated Institution
The shareholding value in the target firm is in the top 10% of their portfolio. Source: WIND database.

TRANINS Transient Institution
The shareholding value in the target firm is not in the top 10% of their portfolio. Source: WIND database.

LONDEDI Long Dedicated Institution
Long monitoring institution indicates the dedicated institutional investors with long-term horizon. Source: WIND database.

SHODEDI Short Dedicated Institution
Short monitoring institution indicates the dedicated institutional investors with short-term horizon. Source: WIND database.

LONTRAN Long Transient Institution
Long monitoring institution indicates the transient institutional investors with long-term horizon. Source: WIND database.

SHOTRAN Short Transient Institution
Short monitoring institution indicates the transient institutional investors with short-term horizon. Source: WIND database.

EARPRE Earnings Pressure
The difference between the consensus of analysts' earnings forecast and firms' potential earnings evaluated at the beginning of the reporter term. Source: CSMAR database.

LONINS Long-incentives Institution
Dummy variables that take the value of 1 if the average holding time is greater than the average position of institutional investors in each industry each year. Source: WIND database.

SHOINS Short-incentives Institution
Dummy variables that take the value of 1 if the average holding time is less than the average position of institutional investors in each industry each year. Source: WIND database.

CASHRA Cash Holding Ratio
The ratio of cash holding over total book assets. Source: CSMAR database.

DUTYMER Duty Merging
Dummy variable that takes the value of 1 if chairman and CEO are the same people; otherwise, the value of 0. Source: CSMAR database.

SEPERATE Separation of Powers
The difference between the ultimate controller's control and cash flow rights. Source: CSMAR database.

BALANCE Equity Balance
The ratio of the shareholding ratio of the second to the tenth largest shareholder to the largest shareholder. Source: CSMAR database. ROAPER Roa Performance The ratio of net income over total book assets. Source: CSMAR database.

BOMARK Book Market Ratio
The ratio of equity book value over equity market value. Source: CSMAR database.

INDEPER Independent Directors Ratio
The ratio of the number of independent directors over the total number of board members. Source: CSMAR database.
EQUNAT Equity Nature Dummy variable that takes the value of 1 if the enterprise is state-owned enterprises, otherwise, the value of 0. Source: CSMAR database.

Appendix B
Following Gaspar et al. [48], this section constructs the proxy of turnover rate (Turnover) in three steps. First, calculating Turnaround indicating the change of the investors' portfolios of investor j and quarter t for two quarters as Turnaround i,t = i∈Q N i, j,t P i,t − N i,j,t−1 P i,t−1 − N i,j,t−1 (P i,t − P i,t−1 ) i∈Q 1 2 (N i,j,t P i,t + N i,j,t−1 P i,−1 ) where Q denotes the set of stocks held by investor j, P i,t is the price of stock i at quarter t, and N i,j,t is the number of stock i held by investor j at quarter t in the reporting term. Second, averaging the four Turnaround annually to calculate the portfolio Turnaround of investors. Third, constructing firm-level Turnaround as the weighted average of the investors' portfolios Turnover.

Appendix C
According to the method of the Jones model [55,56], this section calculates discretionary accruals in two steps. First, running the following cross-sectional model by each year and CSRC industry classification to estimate the parameters α 1 , α 2 , and α 3 where TA i,t equals net income minus cash flow from operations of enterprise i at time t, ∆REV i,t is the changes in sales revenue, PPE i,t is gross property, plant, and equipment, and A i,t−1 are total book assets of time t−1. Second, using the estimated parameters α 1 , α 2, and α 3 to calculated discretionary accruals of firm i at time t as where ∆AR i,t is the change in receivables of firm i at time t. As all variables are scaled, the enterprise's discretionary accruals are indicated as a percentage of the enterprise's total book assets.