CEO's Hometown Identity and Corporate Social Responsibility

Abstract

This study documents the effect of CEO's identification with their hometown on corporate social responsibility (CSR). We propose that firms headquartered in their CEOs’ hometowns tend to do more CSR. This is because identification with their hometown activates CEOs’ altruistic tendency to be more prosocial and makes them more likely to have long-term goals, both of which are compatible with the nature of CSR. This hometown identity effect is stronger when the firm is more locally connected and is weaker when the firm is located in a region with more diverse dialects. Analyzing a large sample of publicly listed Chinese firms for 2009–2016, we found strong support for our predictions. The robustness of our findings is confirmed by a field survey, a difference-in-differences (DID) approach, the Heckman two-stage model, the impact threshold of confounding variables (ITCV), and alternative measures of CSR and CEO hometown identity.

Keywords

CEO hometown identity upper echelons theory CSR local connection dialect diversity

Introduction

Corporate social responsibility (CSR) reflects the extent to which the firm responds actively to demands from a broad range of stakeholders (Carroll, 1979; Mattingly & Berman, 2006; Mcwilliams & Siegel, 2001; Wood, 1991). As general audiences pay increasing attention to firms’ activities in the social domain (Margolis & Walsh, 2003; Mcwilliams & Siegel, 2001), the role of the CEO in firms’ engagement in CSR initiatives has been investigated widely (e.g., Chin, Hambrick, & Treviño, 2013; Petrenko, Aime, Ridge, & Hill, 2016; Tang, Mack, & Chen, 2018; Tang, Qian, Chen, & Shen, 2015).

Exploring the influence of CEOs on CSR is rooted in upper echelons theory that suggests that firms’ strategic decisions and outcomes reflect their top leaders’ personal characteristics (Hambrick & Mason, 1984). A significant proportion of the variance in firms’ CSR engagement can be explained by CEO-level heterogeneity (Wernicke, Sajko, & Boone, 2021). For example, certain cognitive or psychological characteristics of CEOs such as hubris or narcissism have been investigated (Petrenko et al., 2016; Tang et al., 2015, 2018). Nevertheless, the role of CEO's social identity in CSR has been less well studied. This research complements existing work by focusing on a CEO's hometown identity, defined as a strong identification, both conscious and unconscious, with fellow townspeople and a sense of belonging to the hometown community (Ratter & Gee, 2012).

Where the decision-maker was born has been shown to have important implications for organizational decisions and outcomes (e.g., Cohen, Coval, & Malloy, 2011; Hodler & Raschky, 2014; Knight, 2006). We hypothesize that firms headquartered in their CEOs’ hometowns engage in increased CSR. This is because hometown identity triggers the CEOs’ altruistic tendency and make them be more long-term oriented, both of which are consistent with the nature of investment in CSR. Moreover, according to the social identity theory, the extent to which an individual is identified with the group can be affected by some contextual factors (e.g., social norm and linguistic environment) (Ashforth & Mael, 1989), and here we in particular focus on two salient factors in our setting: the firm's local connection and dialect diversity in the region where the firm is headquartered. Local connection strengthens CEOs’ identification with the hometown and so does their prosocial tendency to help fellow townspersons; in contrast, dialect diversity, inherently reflecting cultural heterogeneity, may not engender the same strength of social identification with the hometown. Following these rationales, we predict that the main effect will be strengthened when the firm is more locally connected and weakened when the firm is located in a region with more diverse dialects. Our predictions are supported by analyzing a large sample of Chinese publicly listed firms for 2009–2016. A difference-in-differences (DID) approach helps address potential concerns about endogeneity.

Our study highlights CEO hometown identity and contributes to the existing literature by integrating upper echelons theory with CSR research. First, our study corresponds to a burgeoningly holistic framework in the strategic leadership literature for the drivers of executive behavior (Wowak, Gomez-Mejia, & Steinbach, 2017). This theoretical model suggests that executives’ behavior can be explained parsimoniously in terms of the motivational frame of reference (individualized versus socialized agency) for conceptualizing rewards and the nature of the rewards (pecuniary or nonpecuniary) that motivate executive action. On one hand, in terms of the motivational frame of reference, hometown identity clearly falls in the category of socialized agency as it identifies the CEO socially with fellow townspersons, which goes beyond merely an individual attribute of CEOs; on the other hand, in terms of the nature of the rewards, CSR reflects the motivational factor in the nonpecuniary domain as a firm's CSR performance mostly brings social benefits to the CEO (Tang et al., 2018). In this regard, our investigation into the relationship between CEOs’ hometown identity and CSR directly tests this holistic framework and consequently generates novel insights into the strategic leadership literature.

Second, our study also contributes to social identity theory by integrating this important theoretical perspective with upper echelons theory. Probably one of the most widely studied and established perspectives in the realm of social psychology, social identity theory describes the sociocognitive processes that address individuals’ interactions and behaviors as influenced by the different categories to which they belong (Tajfel & Turner, 1979). Hometown is clearly an important social category with which individual CEOs will self-identify as a member. Meanwhile, upper echelons theory highlights how individual executives’ sociocognitive attributes influence firms (Hambrick & Mason, 1984). However, the integration of social identity theory with upper echelons theory has been rather limited in the literature. In this regard, our study makes contributions to both theoretical perspectives by filling this intellectual void.

Moreover, this study contributes to CSR research by revealing an important yet underexplored antecedent from the perspective of individual CEOs. Where CEOs were born and raised can have a persistent and significant influence on their strategic decisions including those in the social domain, especially for Chinese CEOs in our sample who usually have spent a significant amount of time in their hometowns during their youth. However, this issue has received limited attention. Our findings thus contribute to understanding the drivers of CSR. In addition, our focus on the role of CEOs’ hometown identity in CSR as well as the moderating effects of firms’ local connection and regional dialect diversity point to the need for a more in-depth investigation of the strategic implications of CEOs’ hometown identity for their firms.

Theoretical Background

How CEOs’ Hometown Identity Affects Firms

Both behavioral geography and environmental psychology recognize the role of the “sense of place” in individuals’ decision-making (Nielsen-Pincus, Hall, Force, & Wulfhorst, 2010; Qian & Zhu, 2014). As the “interplay of affect and emotions, knowledge and beliefs, and behaviors and actions in reference to a place” (Altman & Low, 1992: 5), sense of place describes the relationship between people and their spatio-environmental settings (Jorgensen & Stedman, 2001). It is manifested by the bonds between the individual and a place and the affective, cognitive, and behavioral components of these bonds (Altman & Low, 1992: 5).

The hometown is, for many people, a special place that can elicit strong sentiments and affect individuals’ cognition and behavior (Scannell & Gifford, 2010). The special relationship that individuals have with their place of birth evokes hometown identity, which is defined as the sentimental bond linking individuals to their hometown. Hometown identity extends individuals’ sense of identity to their hometown and fellow townspeople. As the place where an individual was born and raised, the hometown affects the early development of cognitive and psychological systems that can have a persistent influence (Moore, 2000). The hometown frequently takes on a special meaning for the individual and can satisfy the need for security, comfort, and consistency (Nielsen-Pincus et al., 2010; Scannell & Gifford, 2010). Individuals invoke their hometown as a social category to establish self-identity, identify with in-group members (fellow townspeople), and define their behavior in a given context (Hogg, Terry, & White, 1995).

According to social identity theory, an individual's self-concept comprises a personal identity (e.g., personality traits; i.e., the “I”) and a social identity (e.g., the group to which the individual belongs; i.e., the “we”) (Ashforth, Harrison, & Corley, 2008). Social identity is based on individuals’ socially defined category of themselves and others (Turner, Hogg, Oakes, Reicher, & Wetherell, 1987). Individuals rely on their social identity to recognize their social environment and define their relationship with others (Ashforth & Mael, 1989). For example, sharing a hometown helps to preserve a distinct Chinese cultural identity for migrant communities both within and beyond China (Fisman, Shi, Wang, & Xu, 2018).

Being a form of social identity, hometown identity continually influences individuals’ cognitive development. Childhood experiences have been shown to have a strong influence on social identity in adult life (Cobb, 1977; Cooper, 1992; Hester & O’ Donnell, 1987). The place where an individual grew up tends to shape the individual's early mental model, and this influence tends to persist for life. The sentimental bond with the hometown allows a strong identification, both conscious and unconscious, with fellow townspeople and creates a sense of belonging to the hometown community (Ratter & Gee, 2012).

This existing research has shown that hometown identity can influence individuals’ decision-making greatly, usually through in-group favoritism that affects resource allocation. For example, Gropper, Jahera, and Park (2013) found that U.S. banks headquartered in the home states of board chairs who are also senators or members of the House of Representatives outperform banks not headquartered in the home states of the board chairs. Using data for 38,427 subnational regions in 126 countries and yearly observations between 1992 and 2009, Hodler and Raschky (2014) found that subnational regions have more intense nighttime lighting if they encompass the hometowns of the incumbent political leaders. Studying fellow selection by the Chinese Academies of Sciences and Engineering, Fisman et al. (2018) found that a hometown shared with selection committee members increases a candidate's chance of selection.

While scholarly evidence of the firm-level implications of CEO hometown identity is limited, anecdotal evidence abounds. For example, Bill Gates moved Microsoft's headquarters to his hometown of Seattle. Phil Knight, the founder of Nike, headquartered his company in his hometown of Portland, where he has spent most of his life. The Chinese e-commerce tycoon Jack Ma not only headquartered the Alibaba Group in his hometown of Hangzhou but also attracted other corporations to the city. Therefore, intellectual curiosity about the influence of a CEO's hometown identity on firm's strategic decisions such as CSR is well deserved.

A valid concern is that the place of birth is not necessarily the place where the CEO grew up. This difference may weaken a CEO's social identification with the hometown. Nevertheless, in the Chinese setting (the sample we use), the place of birth is very likely to be where an individual grows up. This was especially true for the period when our sampled CEOs were young (i.e., before the 1990s). China has adopted a strict “registered residence” (“hukou”, in mandarin Chinese) system, under which interregion mobility (even within the same municipality) was greatly constrained until the late 1980s. Although this system has been loosened recently, the majority of (if not all) Chinese still remain in the same city (or county) before they go to college. For example, in the year 2000, only about 7.9% of the entire population in China were mobile. Since about 99.6% of our sampled CEOs were born before 1980, when these CEOs grew up, they were very likely to remain where they were born, because of the very strict hukou policy.¹

The Role of CEOs in CSR

CSR reflects the extent to which a firm responds actively beyond legal requirements to demands from a broad range of stakeholders (Carroll, 1979; Mattingly & Berman, 2006; Mcwilliams & Siegel, 2001; Wood, 1991). In fact, CSR can be a costly long-term investment without immediate financial returns (Chen, Zhou, & Zhe, 2019; Lorenzo-Molo & Udani, 2013).

As CSR plays a pivotal role in a firm's long-term competitiveness (Hillman & Keim, 2001), keen attention is being devoted to the antecedents of firms’ engagement in CSR. Existing research has shown that CEOs’ personal characteristics substantially predict CSR. For instance, based on a sample of 393 CEOs of S&P 500 companies, Slater and Dixon-Fowler (2009) showed that CEOs with experience of international assignments are linked to more extensive engagement in socially responsible activities by their firms. In a sample of 650 publicly listed U.S. firms, Manner (2010) found that female CEOs with a bachelor's degree in humanities and broad career experience are likely to inspire increased CSR. Some studies have looked at how executive tenure and compensation affect CSR (Deckop, Merriman, & Gupta, 2006; Manner, 2010). For example, among 313 S&P firms, Deckop and colleagues (2006) find that a short-term focus on CEOs’ pay is related negatively to CSR while a long-term focus shows a positive relation to CSR. In another study, Kang (2010) finds that as CEOs approach retirement, their firms pay less attention to social issues and engage in fewer socially responsible initiatives.

This line of investigation has recently been extended to the domain of executives’ cognitive and psychological factors. For example, Chin et al. (2013) find that CEOs with more liberal political ideologies tend to engage more actively in CSR; Tang et al. (2015) show that hubristic CEOs pay less attention to CSR, while Petrenko et al. (2016) demonstrate that narcissistic CEOs engage in more CSR initiatives. Tang et al. (2018) replicate the contrasting effects of CEO hubris and narcissism and further show that the two effects can be moderated by the CSR activities of peer CEOs. Garmache, Neville, Bundy, and Short (2020) demonstrate that CSR initiatives targeting different groups of stakeholders depend on a CEO's regulatory focus (i.e., promotion focus vs. prevention focus). Despite all these insights, a CEO's social identity has been studied less. Social identity should influence the process by which a CEO collects, processes, and interprets information, resulting in different strategic outcomes (Hambrick & Mason, 1984). To fill this intellectual void, this study focuses on a unique form of social identity, hometown identity, and how a CEO's hometown identity can influence CSR.

Theory and Hypotheses

The Effect of CEO Hometown Identity on CSR

We propose that firms headquartered in the CEOs’ hometowns will do more CSR. This is driven by two principal mechanisms. On the one hand, common hometown roots provide an important categorical basis for constructing a shared social identity (Douw, Huang, & Godley, 1999). According to social identity theory, individuals tend to consider as members of the in-group those belonging to the same social category as themselves and as out-group members those from a different social category (Tajfel, 1974; Tajfel & Turner, 1979; Turner, 1982). Individuals tend to have positive emotions towards in-group members and thus treat them more favorably (Ashforth & Mael, 1989). In the hometown context, CEOs tend to consider their fellow townspeople as in-group members and treat them favorably. Such in-group favoritism naturally fosters an altruistic and other-regarding tendency, resulting in prosocial activities.

As a form of prosocial/altruistic behavior (Benabou & Tirole, 2010), CSR allows CEOs to contribute to their hometowns effectively (Borghesi, Houston, & Naranjo, 2012). Hometown triggers a social identification process, during which sentimental drivers prompt CEOs to increase consideration of the interests of “hometown groups.” When the CEO is identified with the CEO's hometown, the CEO is more likely to help his or her hometown, as in regional favoritism by government officials (Hodler & Raschky, 2014). Therefore, CEOs will improve the social welfare of their hometowns and increase CSR participation.

Moreover, as holders of a prestigious position, CEOs may consider themselves as in-group prototypes for their fellow townsfolk. Individuals with higher status can attract greater attention and exert more influence than their peers (Flynn & Amanatullah, 2012). With this consideration, hometown CEOs may feel more obliged to fulfil their fellow townsfolk's expectation and act for their benefit (Hogg & Terry, 2000). Consequently, hometown CEOs will put more emphasis on mutual benefits within the local community and devote more effort to improving local well-being (Tidwell, 2005). Therefore, hometown CEOs tend to do more social good and improve CSR.

In addition, social identity theory predicts that identification with a community increases altruism toward that community (Turner, 1982). Strong identification with a group should promote more activities that benefit the group and discourage those that might harm the group (Boivie, Lange, McDonald, & Westphal, 2011; Dukerich, Golden, & Shortell, 2002). For this reason, hometown identity may motivate the CEO to be more altruistic and prosocial and make decisions to prioritize the interests of hometown stakeholders.

On the other hand, hometown identity may be associated with the CEO's long-term orientation (Uzzell, Pol, & Badenas, 2002). Hometowns are usually attached with old memories that the CEO wishes to retain and doesn’t want to lose through abrupt changes. This desire requires the CEO to preserve the long-term development of the local community. Existing research has shown that identification with a particular region will lead individuals to strive for the sustainability of the region (Uzzell et al., 2002). Moreover, social identity theory points out that identification with the group will drive group members to care more about the common interests of the entire group, such as the development and sustainability of the group (Tidwell, 2005). In our setting, hometown CEOs tend to identify with the place where they were born and grew up, and this identification will spur CEOs to attend to the sustainable development of their hometown as well as their fellow townsfolk's long-term well-being. For example, Lai, Li, and Yang (2020) have shown that if a firm is located in the CEO's hometown, the CEO tends to take long-term actions to meet local stakeholders’ expectations.

CEOs with a long-term orientation emphasize the development of strategic resources that do not generate immediate returns (Hamel & Prahalad, 1989, 1994). They more likely acknowledge the latent benefits of better stakeholder relationships arising from CSR activities. Hometown identity can also trigger a positive attitude toward environmental sustainability (Hernández, Martín, Ruiz, & Hidalgo, 2010) and encourage environmentally friendly behavior (Vaske & Kobrin, 2001). All these issues are consistent with the requirements of CSR, as CSR is by definition a long-term investment with no immediate return. Indeed, CSR is part of a firm's long-term strategy and the returns from CSR investment are not realized in the short run (Crilly, Ni, & Jiang, 2016). Therefore, hometown identity channels CEOs’ attention toward CSR issues.

To sum up, we predict that firms headquartered in their CEOs’ hometowns will engage more actively in CSR. Therefore, we hypothesize that:

Hypothesis 1: Firms headquartered in their CEOs’ hometowns show a higher level of CSR activities.

Moderating Effects

We next explore the boundary conditions of the relationship between CEOs’ hometown identity and CSR. The purpose of this exploration is as follows: While the mechanism caused by a CEO's hometown identity might be difficult to assess directly, we can validate this mechanism by identifying moderators that can strengthen (weaken) this proposed mechanism; in other words, if the main effect becomes stronger (weaker) for the strengthening (weakening) moderator, we can have greater confidence in the proposed mechanism (cf., Tang et al., 2015; Wang & Qian, 2011).

Social identity theory suggests that the extent to which an individual identifies with their social group is critically determined by the contextual factors. We adopt a multilevel approach and focus on the following two contextual factors, firm-level local connection and regional-level dialect diversity, both of which will affect the extent to which CEOs identify with their hometown.

Firm local connection

Local connection measures the extent to which a firm is tied closely to the local community through various interorganizational connections (Lok & Willmott, 2019). The more locally connected a firm is, the more social and economic interactions there are between the firm and other local social entities (Bammens & Hünermund, 2020). According to social identity theory, repeated exchanges between group members help to strengthen individual members’ identification with the group (Hogg et al., 1995). Therefore, a firm's local connection gets stronger as the interaction between the firm and the local community increases (Lai et al., 2020; Mattingly & Berman, 2006). During this process, local connection enhances the firm's prosocial attitude to the local community.

As emphasized by social identity theory, when individuals identify with the group, “a psychological group” emerges, which is “a collection of people who share the same social identification or define themselves in terms of the same social category membership” (Turner, 1985). Although a psychological group need not be interpersonal or based on interaction, as interactions between group members increase, their affinity will be strengthened (Ashforth & Mael, 1989). Therefore, a CEO who is steering a firm that is more locally connected can identify more strongly with the CEO's hometown people and then be more intensely devoted to the collective welfare of the local community.

At the same time, when a firm is more locally connected, the CEO is more likely to receive attention from fellow townspeople, enhancing the CEO's self-perception as a prototype of the hometown community. Therefore, the CEO will be more eager to take care of local interests and the CEO's CSR motivation becomes stronger. Moreover, a more locally connected firm tends to have more ties to local businesses. This will consolidate their common interest with the local community (Lai et al., 2020). For this reason, if CSR can benefit the local community, such benefits should spill over to the firm. Therefore, hometown CEOs are more motivated to engage with CSR. By contrast, for firms that are less locally connected, the stakeholders tend to be outside the hometown. Then CSR may benefit outsider stakeholders rather than local stakeholders. Since hometown CEOs do CSR because they care for local welfare and sustainability, a lower level of firm local connection will weaken a hometown CEO's motivation for CSR. Therefore, we hypothesize that:

Hypothesis 2: The effect of CEO hometown identity on CSR is stronger for firms that are more locally connected.

Regional dialect diversity

Various contextual contingencies such as language, culture, and social norms strongly influence individuals’ beliefs and value systems, and consequently affect how much they acknowledge their social identity (Ashforth & Mael, 1989). Therefore, we investigate whether the diversity of dialects in the region where the firm is headquartered affects the influence of hometown identity on CSR.

Diversity of dialects directly reflects the level of cultural heterogeneity in a region (Frijns, Dodd, & Cimerova, 2016; Qian, 2013). Whether the regional culture is heterogeneous or homogenous critically determines the level of social identification in the region. Language has been considered to be an important cultural element of sense-making as well as sense-giving (Antonsich, 2010). It not only triggers a feeling of belonging to a community but also resonates with a group's feeling of “being home.” Social identity theory suggests that the symbolic meaning attached to a language can shape a unique identity and generate a sharp boundary between groups (Ashforth & Mael, 1989). Indeed, languages can effectively differentiate in-group from out-group members (Antonsich, 2010).

Languages not only define common rules within a group by providing uniqueness but also enhance continuity within a group through sharing symbolic meanings. Therefore, sharing the same languages strengthens individual members’ identification with the groups (Scannell & Gifford, 2010). As Huy (2011) noted, languages work as an identification symbol and help to trigger and maintain collective emotions in a group, channeling individual members’ attention and action toward the group.

Therefore, in regions where local languages are more homogenous, it is much easier to facilitate individual members’ identification with the group, minimizing ingroup differences and maximizing intergroup differences (Huy, 2011; Turner et al., 1987). When CEOs are from a hometown with a single dialect or more similar local dialects, they are more likely to identify with the consensus community, be more embedded in the local group, and enhance their desire for a local identity. All these factors will contribute to a stronger devotion to the local welfare.

In contrast, when the hometown has more diverse dialects, the lack of symbolic coherence will deter the formation of strong identification between group members (Ashforth & Mael, 1989; Hogg et al., 1995). A higher level of dialect diversity will blur the boundary defining each group and constantly remind group members about the cultural heterogeneity. This will create psychological uncertainty in group members regarding self-identification (Hogg et al., 1995). For example, Wright (2011) found as diverse immigrants increase, the common “we feeling” is largely diluted, reducing the level of trust in the local community and threatening identification between local members. Similarly, Hamamura (2017) found that cultural diversity, mainly reflected by linguistic diversity, generates social distance and exclusion; embracing cultural diversity tends to reduce group members’ identification with their group. Therefore, as dialect diversity increases in the hometown, CEOs’ identification with their hometown tends to weaken, reducing the effect of hometown identity on CSR. Therefore, we propose that:

Hypothesis 3: The effect of CEO hometown identity on CSR is weaker for firms located in a region with more diverse dialects.

Method

Sample

We test our hypotheses with a sample of Chinese publicly listed firms that issued CSR reports. Our sample includes firms listed on the Shanghai and Shenzhen Stock Exchanges during 2009–2016. We merged data from the China Stock Market & Accounting Research Database (CSMAR), the Wind Database, corporate annual reports, CSR reports, the National Bureau of Statistics, and the Rankins CSR Ratings (RKS; http://www.rksratings.com). We obtained biographic information for CEOs from corporate annual reports. We manually collected gender, age, educational background, hometown, and other data for the sampled CEOs. We obtained CSR report scores from RKS. Since 2009, RKS has provided CSR scores for all firms listed in China that issued CSR reports. We delete observations with missing values for key variables. We exclude listed firms for which CSR reporting was not available during the sample period and also remove financial firms. The final sample comprises 3,841 firm-year observations.

Measures

Independent variable

CEOs’ hometown identity (Hometown) is measured as follows: If the firm is headquartered in the CEO's hometown, Hometown equals 1 and 0 otherwise. For most of our sampled CEOs, we only have information on hometown at the provincial level, so the main analysis uses CEO hometown at the provincial level.

Dependent variable

We obtain the firm's CSR rating from RKS whose organization and China-specific products are modeled after the U.S. social investment rating agency Kinder, Lydenberg, Domini & Co. Inc. (KLD) (Marquis & Qian, 2014). We use a firm's overall CSR initiatives instead of its CSR activities in the headquarter region (i.e., the CEO's hometown). This is because publicly listed firms’ CSR activities are usually centered within the regions surrounding their headquarters (Guthrie, 2003; Marquis, Glynn, & Davis, 2007; McElroy & Siegfried, 1986; Useem, 1988). For example, studying a sample of firms headquartered in Minneapolis–Saint Paul, Galaskiewicz (1997) find that about 70% of the corporate philanthropy of these firms was within the region of the twin cities.

Like KLD, RKS is independent of the firms it rates. RKS data are taken from firms’ CSR reports and other communications such as websites and press releases. The RKS rating consists of three dimensions: (1) the overall evaluation includes the firm's CSR strategy, the extent of stakeholder participation in CSR activities, comparability of reported information over time, innovativeness of the firm's CSR activities, and the extent of external auditing; (2) the content evaluation focuses on the extent to which leadership and organizational systems are in place to implement CSR, and on specific metrics for economic, environmental, and social responsibilities; and (3) the technical evaluation focuses on items such as the transparency, regularity, and availability of CSR information. The overall rating ranges from 0 to 100; a higher score indicates a higher level of CSR engagement (CSR). RKS data for academic research has been validated by other research into Chinese firms’ CSR (Luo, Wang, & Zhang, 2017; Marquis & Qian, 2014).

Moderating variables

A firm's local connection is measured by the ratio of local to total sales. Local sales data are obtained from the Wind database. The greater the sales from the local market, the more likely it is that the firm is connected with local stakeholders, including employees, consumers, distributers, governments, and communities. Firms’ local connection (Local) takes the value of 1 if the firm's ratio is more than 0.5 and 0 otherwise.

Our second moderator, regional diversity of dialects (Dialect), is measured by the number of dialects in the focal province where the CEO's hometown is. The dialects in our setting include both dialects spoken by the majority Han ethnic group and languages spoken by minority ethnic groups in China. We obtained dialect information in different regions from A Dictionary of Chinese Dialects (Xu & Miyata, 1999) and the Atlas of Chinese Language (Chinese Academy of Social Sciences, & Australian Academy of Humanities, 2012). There are a total of 105 regional dialects in our sample.

Control variables

We control for other variables that may affect CSR at the following three levels: CEOs’ personal characteristics, firm, and region/industry.

For control variables at the CEO level, we control for CEO Gender, which is 1 if the CEO is male and 0 otherwise. We control for CEO age because a CEO's opinion about the importance of CSR can vary with age (Petrenko et al., 2016). CEO Tenure is measured by the number of years since the CEO assumed the position. CEO Education is measured as 1 = secondary school and below, 2 = college, 3 = undergraduate degree, 4 = postgraduate degree, 5 = doctorate degree, and 6 = other (e.g., honorary doctorate degree). Dual is coded as 1 if the CEO is also the chairman of the board and 0 otherwise. CEO Compensation is measured by the natural logarithm of the sum of CEO salary, bonus, and values of stock option and restricted stock. CEO International is a dummy variable that is set to 1 if the CEO has overseas experience (such as work or education), and 0 otherwise. We control for a CEO's political connections (Scpolitical) with a dummy variable that is coded as 1 if the CEO is a member of the National or Provincial People's Congresses or Political Consultative Conference and 0 otherwise. We control for a CEO's involvement in other social activities (Scother) with a dummy variable that is set to 1 when a CEO is a member of a trade association, charity, scientific institution, or other nonprofit organization (Faleye, Kovacs, & Venkateswaran, 2014) and 0 otherwise. We also control for a CEO's directorship in other listed firms (Scfirm) with a dummy variable coded as 1 if the CEO serves as a director for other listed firms and 0 otherwise.

We control for a set of firm characteristics. Firm age is measured as the number of years since the date the firm was first listed on a Chinese stock exchange. Size is the natural logarithm of total assets. We control for firms’ financial performance as captured by the firm's return on assets (ROA) (Adams & Hardwick, 1998; Waddock & Graves, 1997). Ownership is coded as 1 if a firm is a state-owned enterprise and 0 otherwise. Top_holder is computed as the ratio of the largest shareholding to the total number of shares. Insti_holder is measured as the proportion of shares held by institutional shareholders. Board size is measured as the count of board of directors. Independence is measured as the percentage of board members who are independent directors. Diversity measures the gender diversity of board directors by a Blau index of $H = 1 - \sum_{i}^{k} = I p_{i}^{2}$ , where I refers to the number of genders of the directors in a board, and p_i refers to the percentage of directors of different genders in the board (Blau, 1977). Leverage is measured as total liabilities divided by total assets. Revenue growth is measured as the revenue increment (ending business income – opening business income) divided by the opening business income. Current ratio is measured as the ratio of current assets to current liabilities. We include a dummy variable of family set to 1 if the firm's largest shareholder or actual controller is a family member and 0 otherwise (Maury, 2006). We also add one dummy variable of founder set to 1 if the CEO was the founder of the firm and 0 otherwise.

We also control for two regional- and industry-level characteristics. Per_GDP is measured as the ratio of GDP to the total population in the focal province. Industry competition (HHI) is measured by the Herfindahl–Hirschman index. Specifically, we use the following formula: $H H I = \sum_{i = 1}^{N} {(\frac{X i}{X})}^{2}$ , where N is the number of firms in an industry, $X_{i}$ is the sales of firm i in a given year, and X is total market sales in that year. Finally, we control for the mean CSR level of all the firms in the same province in the prior year (Pre_CSR) to capture any unobservable regional-level influence (Li & Lu, 2020).

Estimation Model

To accommodate the nonindependent nature of our longitudinal sample, we analyzed our data using generalized estimating equations (GEEs). GEEs help account for unobserved differences across CEOs as well as intertemporal correlations among outcome variables for individual CEOs (Liang & Zeger, 1986; Lipsitz, Fitzmaurice, Orav, & Laird, 1994) and have been used in prior studies of CEO attributes (e.g., Wowak, Mannor, Arrfelt, & McNamara, 2016). GEEs require specifying the distribution of the dependent variable, the link function, and the intertemporal correlation structure (Aguinis, Gottfredson, & Joo, 2013). We specified a Gaussian distribution with an identity link function. In all models, we specified an exchangeable correlation structure (grouped by CEO) and robust standard errors, which account for any misspecification in the correlation structure (White, 1980).

Results

Table 1 presents the descriptive statistics and correlations of our variables. Hometown CEOs account for 52% of the full sample. The mean score of CSR is 39.16. The variance inflation factor (VIF) scores are all less than 10, with a mean of 2.01 and a maximum of 7.75, suggesting that multicollinearity may not be a serious issue in this study (O'Brien, 2007).

Table 1

Descriptive Statistics and Correlations

	（1）	（2）	（3）	（4）	（5）	（6）	(7)	（8）	（9）	（10）	（11）	（12）	（13）	（14）	（15）	（16）	（17）	（18）	（19）	（20）	（21）	（22）	（23）	(24)	（25）	（26）	(27)	(28)	(29)	(30)	(31)	(32)
(1) CSR
(2) Hometown	.12
(3) Local	.00	–.01
(4) Dialect	–.05	.12	–.01
(5) Lambda	.21	.01	–.06	–.09
(6) Gender	–.02	–.02	–.01	.01	–.03
(7) CEO age	.13	–.02	–.02	–.06	–.11	.05
(8) Tenure	.03	.16	–.01	.04	–.10	.01	.22
(9) Education	.06	–.12	.08	–.11	–.10	–.01	–.07	–.15
(10) Dual	–.04	.03	–.02	.04	.11	.01	.16	.12	–.02
(11) Compensation	–.03	.05	.00	.08	–.03	–.05	–.04	.13	–.04	.02
(12) International	.04	–.04	–.09	–.04	.03	.00	.05	.04	.01	.04	–.03
(13) Scpolitical	.03	.13	.04	.09	.01	–.02	.11	.18	–.05	.27	.01	.05
(14) Scother	–.01	.01	–.05	.04	.08	–.03	–.01	.09	–.02	.30	.02	.03	.14
(15) Scfirm	.09	.07	–.02	.04	–.01	.03	.05	.13	–.08	.18	–.03	.06	.13	.10
(16) Family	–.13	.05	–.10	.14	.34	–.09	–.16	.03	–.17	.24	.06	.09	.09	.21	.09
(17) Founder	–.03	.00	–.02	.01	.16	.04	.08	.02	.01	.43	.01	.04	.18	.28	.11	.28
(18) Firm age	.04	.04	.14	.06	–.04	–.04	.07	.14	.12	–.02	.05	–.01	.02	–.02	–.10	–.09	–.06
(19) Size	.46	–.05	–.03	–.13	–.50	.03	.17	.01	.11	–.11	–.02	.02	.02	–.07	.07	–.34	–.12	.07
(20) Ownership	.13	–.07	.13	–.16	–.31	.07	.15	–.07	.17	–.27	–.08	–.07	–.10	–.23	–.08	–.33	–.25	.05	.32
(21) Top_holder	.14	–.08	–.01	–.19	–.13	.05	.13	–.16	.10	–.12	–.12	–.03	–.02	–.11	.05	–.25	.02	–.22	.29	.30
(22) Insti_holder	.12	.04	.05	–.02	–.07	–.03	.08	.01	.05	.03	–.01	–.03	.01	.01	.00	–.03	.01	.17	.10	.02	.00
(23) Board size	.16	.04	.04	.05	–.19	.08	–.02	.05	.06	–.13	–.02	–.04	–.03	–.09	.04	–.22	–.09	.01	.25	.22	.05	.06
(24) Independence	.05	–.03	–.02	–.10	–.04	–.02	.05	–.02	.01	.09	–.03	–.01	.07	.07	.06	.01	.06	–.08	.11	–.02	.09	–.02	–.28
(25) Diversity	–.01	.03	.07	–.03	.05	–.12	–.04	–.02	–.01	.01	.00	.01	–.01	–.02	–.02	.10	.06	.03	–.06	–.09	–.02	.05	.07	–.04
(26) ROA	.00	.12	–.05	.03	.00	–.05	–.01	.07	–.05	.08	.12	–.01	.10	.04	.05	.15	.05	–.06	–.09	–.17	.02	.11	.00	–.03	.01
(27) Leverage	.13	–.09	.07	–.04	–.21	.01	.02	–.04	.08	–.11	–.03	–.01	–.03	–.04	–.01	–.22	–.10	.12	.51	.24	.09	–.02	.12	.03	–.04	–.44
(28) Revenue growth	–.03	.00	–.01	.07	.03	–.02	–.10	–.03	–.02	.07	.08	.00	.04	.06	.02	.10	.08	–.07	.00	–.10	–.09	.01	.01	.01	.01	.25	.03
(29) Current ratio	–.12	.02	–.06	.01	.20	–.02	.01	.00	–.03	.15	.05	.01	.04	.03	.02	.21	.10	–.10	–.36	–.21	–.09	.01	–.15	.00	.03	.26	–.63	.02
(30) Per_GDP	.24	–.08	.06	–.52	–.04	.03	.11	.04	.09	.06	.01	.05	–.04	.05	.01	.01	.06	.06	.14	.00	.10	.15	–.10	.07	.03	.00	–.03	–.05	.06
(31) HHI	.07	–.02	–.03	–.08	–.12	.01	.07	.02	–.07	–.07	–.09	.08	.06	–.07	–.04	–.09	–.03	–.11	.16	.11	.14	–.07	.11	.01	.01	–.07	–.01	–.06	–.09	–.02
(32) Pre_CSR	.42	–.04	.06	–.07	–.03	–.04	.15	.10	.12	.04	.00	.02	–.02	–.01	.02	–.07	.03	.27	.24	.06	.04	.23	.01	.06	.01	–.06	.02	–.09	.01	.51	-.01
Mean	39.16	.52	.19	7.79	0.58	.95	49.53	4.47	4.08	.16	12.84	.06	.18	.05	.53	.31	.03	16.42	23.02	.64	.39	.24	9.32	.37	.03	.04	.50	.11	1.87	10.93	.11	37.53
SD	11.46	.50	.39	4.51	0.66	.22	5.97	3.44	1.19	.37	2.70	.24	.39	.22	.50	.46	.18	5.93	1.44	.48	.16	.23	2.02	.06	.17	.05	.20	.25	1.82	.45	.11	5.54

Note: N = 3841. Coefficients with a magnitude greater than 0.03 are significant at the .05 level.

Table 2 presents the GEE results of the regression analysis on the effect of CEO hometown identity on CSR. Model 1 includes all the control variables, and Models 2–6 include the independent variable, the moderating variables, and the interaction terms. Hypothesis 1 predicts that firms headquartered in their CEOs’ hometowns will do more CSR. In Model 2, the coefficient of Hometown is significantly positive (β = .71, SE = .36, t = 1.96, ρ = .051), supporting Hypothesis 1. This result is also economically significant: For those firms that have a hometown CEO, the CSR increases significantly by 0.71 points, about 1.81% of the sample mean (39.16).

Table 2

GEE Results of CEO Hometown Identity and CSR

Variable	CSR	CSR	CSR	CSR	CSR	CSR
Variable	(1)	(2)	(3)	(4)	(5)	(6)
Gender	−0.55	−0.41	−0.41	0.46	−0.35	−1.45
	(0.85)	(0.86)	(0.86)	(0.60)	(0.84)	(1.42)
CEO age	0.00	0.00	0.00	−0.01	−0.00	0.04
	(0.03)	(0.03)	(0.03)	(0.03)	(0.03)	(0.05)
Tenure	0.02	0.02	0.02	−0.04	0.02	−0.08
	(0.05)	(0.05)	(0.05)	(0.04)	(0.05)	(0.08)
Education	0.16	0.18	0.18	0.02	0.19	0.03
	(0.16)	(0.16)	(0.16)	(0.13)	(0.16)	(0.26)
Dual	−0.76	−0.75	−0.75	−0.62*	−0.75	−1.10
	(0.51)	(0.52)	(0.52)	(0.36)	(0.51)	(0.71)
Compensation	0.10**	0.09**	0.10**	0.02	0.10**	−0.07
	(0.04)	(0.04)	(0.04)	(0.03)	(0.04)	(0.09)
International	1.37	1.41	1.41	0.54	1.44	2.20**
	(0.92)	(0.92)	(0.92)	(0.61)	(0.91)	(1.10)
Scpolitical	0.74	0.66	0.66	1.13**	0.71	0.30
	(0.57)	(0.58)	(0.58)	(0.56)	(0.57)	(0.86)
Scother	−0.14	−0.26	−0.26	−1.53	−0.23	1.08
	(0.98)	(0.96)	(0.96)	(0.99)	(0.93)	(1.03)
Scfirm	0.39	0.38	0.39	0.46**	0.40	0.41
	(0.28)	(0.27)	(0.27)	(0.20)	(0.27)	(0.50)
Family	−0.51	−0.39	−0.39	−0.91	−0.44	1.28
	(0.94)	(0.93)	(0.93)	(1.01)	(0.94)	(1.77)
Founder	0.50	0.34	0.34	0.58	0.24	0.06
	(1.19)	(1.19)	(1.19)	(1.16)	(1.20)	(1.35)
Firm age	−0.13**	−0.13**	−0.13**	−0.14**	−0.13**	−0.16***
	(0.06)	(0.06)	(0.06)	(0.06)	(0.06)	(0.06)
Size	2.44***	2.44***	2.44***	2.25***	2.44***	3.50***
	(0.28)	(0.28)	(0.28)	(0.29)	(0.28)	(0.37)
Ownership	−0.02	0.05	0.06	−0.55	−0.01	1.20
	(0.98)	(0.97)	(0.97)	(1.10)	(0.98)	(1.80)
Top_holder	−0.63	−0.69	−0.69	0.50	−0.57	−0.14
	(1.70)	(1.70)	(1.70)	(1.46)	(1.70)	(2.47)
Insti_holder	1.15	1.15	1.14	0.03	1.16	1.26
	(0.74)	(0.74)	(0.74)	(0.50)	(0.74)	(1.39)
Board size	0.22*	0.22*	0.21*	0.07	0.21*	0.38**
	(0.12)	(0.12)	(0.12)	(0.10)	(0.12)	(0.19)
Independence	0.27	0.17	0.17	−0.58	0.30	1.60
	(3.00)	(3.00)	(3.00)	(2.65)	(2.98)	(4.52)
Diversity	0.44	0.44	0.44	1.08*	0.46	−0.22
	(0.94)	(0.94)	(0.94)	(0.57)	(0.94)	(1.34)
ROA	4.54	4.53	4.51	−1.60	4.53	−5.58
	(3.63)	(3.64)	(3.64)	(2.53)	(3.62)	(5.35)
Leverage	−0.95	−0.88	−0.87	−1.14	−0.82	−4.56*
	(1.38)	(1.38)	(1.38)	(1.09)	(1.38)	(2.40)
Revenue growth	−0.18	−0.18	−0.18	−0.09	−0.16	0.32
	(0.36)	(0.36)	(0.36)	(0.30)	(0.36)	(0.60)
Current ratio	−0.09	−0.09	−0.09	−0.13*	−0.09	−0.12
	(0.08)	(0.08)	(0.08)	(0.07)	(0.08)	(0.16)
Per_GDP	−2.66	−2.65	−2.65	−2.33	−2.54	−1.87
	(2.30)	(2.31)	(2.31)	(2.32)	(2.33)	(3.08)
HHI	−5.04	−5.22	−5.20	1.93	−5.01	−4.15
	(5.08)	(5.04)	(5.04)	(4.01)	(5.04)	(5.54)
Pre_CSR	0.26***	0.26***	0.26***	0.04	0.25***	0.40***
	(0.06)	(0.06)	(0.06)	(0.05)	(0.06)	(0.08)
Hometown		0.71*	0.71*	0.13	0.72**	2.93***
		(0.36)	(0.36)	(0.30)	(0.36)	(0.64)
Local			−0.25	−0.77	−0.25	−0.86
			(0.83)	(0.99)	(0.83)	(1.13)
Dialect			−0.36	−0.19	−0.29	−0.11
			(0.37)	(0.40)	(0.38)	(0.46)
Hometown * Local				1.37**		2.37*
				(0.60)		(1.36)
Hometown * Dialect					−0.65*	−1.07*
					(0.37)	(0.61)
Year fixed	Y	Y	Y	Y	Y	Y
Industry fixed	Y	Y	Y	Y	Y	Y
Province fixed	Y	Y	Y	Y	Y	Y
Constant	−2.97	−3.35	−2.09	5.99	−3.76	−41.29
	(26.92)	(27.06)	(27.95)	(28.05)	(28.24)	(37.80)
Observations	3,841	3,841	3,841	3,437	3,841	3,841
Wald Chi²	9868.73***	9945.76***	9226.28***	48767.76***	15253.27***	7619.67***

Note: Standard errors are clustered at the firm level.

*p < .10. **p < .05. ***p < .01 (two-tailed tests).

Models 4–6 test Hypotheses 2 and 3. Model 4 includes the interaction between CEO hometown identity and local connection to test Hypothesis 2. The interaction term is significantly positive (β = 1.37, SE = .60, t = 2.30, ρ = .022), showing that local business connections strengthen the positive effect of hometown identity on CSR. Figure 1 illustrates the moderating effect of local connection: When the local connection increases, the positive relationship between CEO hometown identity and CSR gets much steeper. Specifically, by setting all the control variables to their mean, CEO hometown identity increases CSR from 38.10 to 38.23 when local connection is 0; however, when local connection is 1, CEO hometown identity increases CSR from 37.33 to 38.83. As the mean CSR in our sample was 39.16, the increase in CSR changes from 0.33% to 3.83%. These results support Hypothesis 2.

Figure 1

Interaction between CEO Hometown Identity and Local Connection

Model 5 adds the interaction term between CEO hometown identity and regional dialect diversity to test Hypothesis 3, which predicts that the relationship between CEO hometown identity and CSR is weaker when the local dialect diversity is high. The coefficient of the interaction term is significantly negative (β = –.65, SE = .37, t = –1.74, ρ = .081), suggesting that diversity of dialects weakens the relationship between CEO hometown identity and CSR. Figure 2 illustrates the moderating effect of diversity of dialects. Specifically, by setting all the control variables to their mean, CEO hometown identity increases the CSR value from 37.66 to 41.31 when dialect diversity is low (1 SD below the mean); however, when dialect diversity is high (1 SD above the mean), CEO hometown identity decreases the CSR value from 37.98 to 35.77. In other words, the increase in CSR changes from 9.32% to −5.64%. Hypothesis 3 is thus supported.

Figure 2

Interaction between CEO Hometown Identity and Regional Dialect Diversity

Model 6 as the full model produces consistent results.

Supplementary Analyses

To show the robustness of our findings, we have conducted multiple supplementary analyses, including a field survey, a DID analysis, the Heckman two-stage model, the impact threshold of confounding variables (ITCV), the alternative measures of CSR and CEO hometown identity, and alternative estimation models.

Field survey

Our main analyses are based on archival data. We develop our mechanisms for the positive relationship between CEO hometown identity and CSR by grounding our arguments in existing theory, without assessing the proposed mechanisms directly. We have proposed that altruistic or prosocial tendencies and long-term orientation motivate a hometown CEO to do more CSR, but the archival data doesn’t allow us to assess a CEO's altruistic or prosocial tendencies and long-term orientation. Therefore, a field survey of a group of real CEOs may help to validate our theories concerning the findings.

We commissioned two established survey companies in China, Credamo and Weidiaocha, to conduct this survey. Each of the two survey companies sent out 300 questionnaires to a randomly selected sample of CEOs across multiple industries and multiple provinces. Before sending out the questionnaires, we carefully checked the sampled firms’ characteristics, such as industry, location, and size. For example, we required that no more than 20% of the sample firms are from the same province and industry. The sampled firms were required to have more than 50 employees, and more than 75% of the sampled firms have more than 100 employees. The respondents were assured that the data collected will only be used for research purposes. We also provided the respondent a modest honorarium for completing the survey. Credamo received 212 valid responses and Weidiaocha 198. Therefore, in total we got 410 responses, and the response rate was 68.3%. Among all the respondents, 58% are male, 63% are founder CEOs, and 75.8% headquarter their firms in their hometown.

To avoid the common method bias (Podsakoff, MacKenzie, Lee, & Podsakoff, 2003), the survey was conducted at two times. The respondents completed the first part (Time 1) of the questionnaire by the last week of September 2021 and completed the second part (Time 2) of the questionnaire by the last week of October 2021. Time 1 questionnaire includes items on hometown identity, long-term orientation, altruism orientation, and demographic information. Time 2 questionnaire includes items on firm local connection, regional dialect diversity, and CSR. A Harman's one-factor analysis further suggests that the common method bias may not be a serious concern in this survey data. Appendix 1 displays the measurement items used in this field survey. As shown, the measures also passed the validity and reliability test well. Appendix 2 displays the descriptive statistics and correlations of the variables.

The analysis with this survey data as shown in Table 3 largely confirms what we found in the main analyses. We also employed the Sobel (1982) test and bootstrapping technique to confirm that the (partial) mediating effects were significant (Preacher & Hayes, 2008). The Sobel test also confirms that the indirect effect of altruistic orientation was statistically different from zero (Z = 7.73, p < 0.001). Although the parameters used from our structural model for the Sobel test take the covariation of the other mediators into account, the Sobel test assumes multivariate normality, which is frequently violated. Therefore, we also generated, by bootstrapping analysis, a 95% “bias-corrected” confidence interval for the specific indirect effect (Preacher & Hayes, 2008). This confidence interval was .076 to .224, demonstrating a statistically significant indirect effect for altruistic orientation. Similarly, we found, using the Sobel test, the indirect effect of long-term orientation to be statistically different from zero (Z = 4.29, p < 0.001), and the 95% confidence interval, derived from bootstrapping, was .344 to .708. Collectively, the analysis with the field survey data largely confirms the findings of the main analysis and importantly helps to validate our proposed mechanism.

Table 3

Results of Field Survey Data

	(1)	(2)	(3)	(4)	(5)	(6)	(7)
Variables	CSR	CSR	CSR	LOT	AO	CSR	CSR
Gender	0.02	0.01	0.01	−0.06	−0.04	0.03	0.03
	(0.05)	(0.05)	(0.05)	(0.07)	(0.05)	(0.05)	(0.05)
CEO age	0.01	0.01	0.02	−0.14**	−0.03	0.02	0.04
	(0.04)	(0.04)	(0.04)	(0.06)	(0.04)	(0.04)	(0.04)
Education	0.01	0.02	0.02	0.12**	−0.01	0.01	−0.02
	(0.04)	(0.04)	(0.04)	(0.06)	(0.04)	(0.04)	(0.04)
Founder	0.14**	0.13**	0.12**	0.14*	0.06	0.11**	0.11*
	(0.06)	(0.06)	(0.06)	(0.08)	(0.06)	(0.05)	(0.06)
Size	0.05	0.05	0.06	0.01	−0.01	0.05	0.04
	(0.04)	(0.04)	(0.04)	(0.05)	(0.04)	(0.04)	(0.048)
Firm age	−0.03	−0.05	−0.05	0.17***	0.01	−0.04	−0.07*
	(0.04)	(0.04)	(0.04)	(0.05)	(0.04)	(0.04)	(0.04)
Industry	−0.04*	−0.04*	−0.04*	−0.10***	−0.05**	−0.02	−0.02
	(0.02)	(0.02)	(0.02)	(0.03)	(0.02)	(0.02)	(0.02)
Region	0.05	−0.01	−0.01	0.03	0.05**	−0.02	0.00
	(0.02)	(0.02)	(0.02)	(0.03)	(0.02)	(0.02)	(0.02)
Headquarters	0.14**	0.12*	0.11*	0.15*	0.02	0.13**	0.11*
	(0.06)	(0.06)	(0.06)	(0.08)	(0.06)	(0.06)	(0.06)
Ownership	0.05	0.05	0.04	−0.10	−0.03	0.06	0.07
	(0.05)	(0.05)	(0.05)	(0.07)	(0.05)	(0.05)	(0.05)
Family	0.02	0.01	−0.01	−0.05	−0.08	0.06	0.03
	(0.06)	(0.06)	(0.06)	(0.08)	(0.06)	(0.05)	(0.06)
Hometown	0.41***	0.35***	0.35***	0.50***	0.68***	0.10*	0.31***
	(0.05)	(0.05)	(0.05)	(0.06)	(0.05)	(0.05)	(0.05)
Local		0.15***	0.12***
		(0.03)	(0.04)
Dialect		0.01	0.02
		(0.02)	(0.02)
Hometown * Local			0.08**
			(0.04)
Hometown * Dialect			−0.07*
			(0.04)
AO						0.47***
						(0.05)
LTO							0.20***
							(0.04)
Constant	3.09***	2.71***	2.76***	2.77***	2.27***	2.03***	2.53***
	(0.36)	(0.37)	(0.37)	(0.48)	(0.34)	(0.34)	(0.36)
Observations	410	410	410	410	410	410	410
R-squared	0.2317	0.2666	0.2806	0.2626	0.4074	0.3797	0.2867

Note: Standard errors are clustered at the firm level.

*p < .10. **p < .05. ***p < .01 (two-tailed tests).

Heckman two-stage model

Since our sample only covers firms that have CSR scores (or have issued CSR reports) during 2009–2016, there might be a selection bias as firms may only issue a CSR report when they have performed well socially. Therefore, we conducted a Heckman two-stage model (Certo, Busenbark, Woo, & Semadeni, 2016; Heckman, 1979; Hill, Johnson, Greco, O'Boyle, & Walter, 2020; Semadeni, Withers, & Certo, 2014).

In the first stage, we conducted a probit model and used dual, CEO international, CEO political connection (Scpolitical), firm size (Size), firm age, ownership, board size, board independence, ROA, leverage, current ratio, Per_GDP, industry competition (HHI), year-, industry-, and region-fixed effects to predict whether a firm issued a CSR report in a particular year. The instrumental variable we used is whether the firm has been required to issue a CSR report (Mcsr). The Stock Exchange Act in China has explicitly required certain types of firms to issue CSR reports. For instance, since May of 2008, the Shanghai Stock Exchange has requested all firms listed in the NYSE and HKSE as well as financial firms to issue CSR reports; the Shenzhen Stock Exchange requires firms included in the SZSE 100 index to issue CSR reports. Therefore, we created a dummy variable that is set to 1 if the firm has been required to issue CSR reports, and 0 otherwise.

We generated an inverse Mills’ ratio (Lambda) from the first stage model and included it in the second stage model. The results in Table 4 are consistent with those in the main analyses.

Table 4

The Second Stage Results of Heckman Model

Variable	CSR	CSR	CSR	CSR	CSR	CSR
Variable	(1)	(2)	(3)	(4)	(5)	(6)
Gender	−1.50	−1.44	−1.44	−1.54	−1.32	−1.42
	(1.51)	(1.42)	(1.42)	(1.43)	(1.41)	(1.42)
CEO age	0.03	0.04	0.04	0.04	0.03	0.03
	(0.05)	(0.05)	(0.05)	(0.05)	(0.05)	(0.05)
Tenure	0.01	−0.03	−0.03	−0.04	−0.03	−0.03
	(0.08)	(0.09)	(0.09)	(0.08)	(0.08)	(0.08)
Education	−0.02	0.08	0.08	0.09	0.06	0.08
	(0.26)	(0.26)	(0.26)	(0.26)	(0.25)	(0.26)
Dual	−1.37*	−1.24*	−1.25*	−1.22*	−1.13	−1.10
	(0.72)	(0.71)	(0.71)	(0.71)	(0.71)	(0.70)
Compensation	−0.03	−0.04	−0.04	−0.05	−0.05	−0.05
	(0.09)	(0.09)	(0.09)	(0.09)	(0.09)	(0.09)
International	1.68	1.86*	1.88*	1.89*	1.97*	1.98*
	(1.13)	(1.11)	(1.11)	(1.12)	(1.09)	(1.09)
Scpolitical	0.83	0.34	0.34	0.32	0.33	0.30
	(0.86)	(0.86)	(0.86)	(0.85)	(0.86)	(0.86)
Scother	0.70	1.02	1.03	0.97	1.02	0.96
	(1.09)	(1.02)	(1.02)	(1.02)	(1.03)	(1.03)
Scfirm	0.51	0.37	0.36	0.35	0.42	0.40
	(0.51)	(0.50)	(0.50)	(0.49)	(0.49)	(0.49)
Family	0.45	0.55	0.55	0.58	0.75	0.77
	(1.84)	(1.76)	(1.75)	(1.75)	(1.76)	(1.76)
Founder	0.05	−0.09	−0.09	−0.18	−0.17	−0.26
	(1.45)	(1.41)	(1.41)	(1.38)	(1.38)	(1.35)
Firm age	−0.15***	−0.16***	−0.16***	−0.16***	−0.16***	−0.16***
	(0.6)	(0.06)	(0.06)	(0.06)	(0.06)	(0.06)
Size	4.06***	4.13***	4.14***	4.12***	4.14***	4.12***
	(0.43)	(0.42)	(0.42)	(0.42)	(0.42)	(0.42)
Ownership	1.19	1.22	1.20	1.24	1.41	1.45
	(1.90)	(1.82)	(1.81)	(1.81)	(1.82)	(1.82)
Top_holder	−1.25	−0.49	−0.48	−0.59	−0.60	−0.71
	(2.49)	(2.48)	(2.48)	(2.49)	(2.47)	(2.48)
Insti_holder	2.02	1.65	1.64	1.59	1.62	1.56
	(1.44)	(1.38)	(1.38)	(1.38)	(1.37)	(1.37)
Board size	0.41**	0.41**	0.41**	0.41**	0.40**	0.40**
	(0.19)	(0.19)	(0.19)	(0.19)	(0.19)	(0.19)
Independence	1.88	1.51	1.50	1.32	1.66	1.49
	(4.59)	(4.48)	(4.47)	(4.46)	(4.47)	(4.47)
Diversity	0.09	−0.23	−0.24	−0.36	−0.39	−0.52
	(1.35)	(1.34)	(1.34)	(1.33)	(1.35)	(1.35)
ROA	−1.04	−3.46	−3.47	−3.22	−3.86	−3.62
	(5.39)	(5.32)	(5.32)	(5.31)	(5.29)	(5.28)
Leverage	−5.04**	−5.11**	−5.13**	−5.26**	−5.10**	−5.23**
	(2.44)	(2.38)	(2.38)	(2.37)	(2.37)	(2.36)
Revenue growth	−0.08	−0.12	−0.11	−0.18	−0.06	−0.13
	(0.62)	(0.61)	(0.61)	(0.61)	(0.61)	(0.62)
Current ratio	−0.20	−0.20	−0.20	−0.20	−0.20	−0.20
	(0.16)	(0.16)	(0.16)	(0.16)	(0.16)	(0.16)
Per_GDP	−1.39	−1.59	−1.59	−1.77	−1.62	−1.80
	(3.01)	(3.04)	(3.04)	(3.04)	(3.07)	(3.06)
HHI	−3.42	−4.29	−4.35	−4.53	−4.15	−4.33
	(5.55)	(5.41)	(5.42)	(5.47)	(5.43)	(5.48)
Pre_CSR	0.45***	0.43***	0.43***	0.42***	0.42***	0.42***
	(0.08)	(0.09)	(0.09)	(0.08)	(0.09)	(0.08)
Lambda	2.18***	2.37***	2.36***	2.35***	2.40***	2.39***
	(0.52)	(0.50)	(0.50)	(0.51)	(0.50)	(0.50)
Hometown		3.52***	3.52***	3.09***	3.53***	3.10***
		(0.57)	(0.57)	(0.64)	(0.57)	(0.64)
Local			0.20	−0.92	0.13	−0.98
			(1.08)	(1.15)	(1.07)	(1.14)
Dialect			−0.12	−0.13	−0.02	−0.03
			(0.44)	(0.44)	(0.45)	(0.45)
Hometown * Local				2.28*		2.26*
				(1.36)		(1.36)
Hometown * Dialect					−1.18*	−1.17*
					(0.61)	(0.61)
Year fixed	Y	Y	Y	Y	Y	Y
Industry fixed	Y	Y	Y	Y	Y	Y
Province fixed	Y	Y	Y	Y	Y	Y
Constant	−61.76*	−60.53*	−60.44	−57.62	−60.92	−58.11
	(36.21)	(36.64)	(37.87)	(37.79)	(38.20)	(38.12)
Observations	3,841	3,841	3,841	3,841	3,841	3,841
R-squared	0.4383	0.4567	0.4567	0.4581	0.4588	0.4601

Note: Standard errors are clustered at the firm level.

*p < .10. **p < .05. ***p < .01 (two-tailed tests).

DID method

To address the endogeneity issue, we employ a DID approach (Huang & Kisgen, 2013). The advantage of the DID model is that it corrects for other (time-invariant unobserved) factors that might be associated with a change in CSR due to a change of CEO. Specifically, we construct the following DID model to test the impact of CEO hometown identity on CSR:

C S R_{i t} = β_{0} + β_{1} T u r n o v e r_{i t} \times P o s t_{i t} + γ X_{i t} + Y e a r_{t} + I n d u s t r y_{c} + P r o v i n c e_{j} + ε_{i t}

(1)

where

T u r n o v e r_{i t}

is set to 1 if the firm is in the treatment group and 0 otherwise.

P o s t_{i t}

is set to 1 if the CEO is replaced in the year t and 0 otherwise. We select a time window of 2–4 years around the CEO change. In the first case, if CEO hometown is 0 before or after a change, the firm joins the control group; if the former CEO was a nonhometown CEO and the successor is a hometown CEO, the firm joins the treatment group. In the second case, if CEO hometown is 1 before/after a change of CEO, the firm joins the control group; if the former CEO was a hometown CEO and the successor is not, the firm joins the treatment group.

In Table 5, Models 1 and 2 present the result of the first case above. In Model 2, the coefficient of Turnover1*Post1 is significantly positive (β = 2.84, SE = .91, t = 3.11, ρ = .002), indicating that a firm's CSR increases significantly if the former CEO was not a hometown CEO but the present CEO is. In Table 5, Models 3 and 4 present the results of the second case above. The coefficient of Turnover2*Post2 is significantly negative (β = –1.86, SE = 1.05, t = -1.77, ρ = .079) in Model 4, indicating that CSR decreases significantly if the former CEO was a hometown CEO but the new CEO is a nonhometown CEO. These DID regression results confirm that CEO hometown identity has a significant positive effect on CSR.

Table 5

DID Analysis

Variable	CSR	CSR	CSR	CSR
Variable	(1)	(2)	(3)	(4)
Gender	1.56	2.32	−4.29*	−4.16*
	(1.73)	(1.84)	(2.29)	(2.26)
CEO age	0.09	0.10	0.15*	0.16*
	(0.07)	(0.07)	(0.09)	(0.09)
Tenure	−0.21**	−0.14	0.23	0.15
	(0.10)	(0.10)	(0.16)	(0.17)
Education	0.05	−0.01	0.62	0.65
	(0.37)	(0.37)	(0.47)	(0.46)
Dual	0.89	0.62	−3.17*	−3.02*
	(1.21)	(1.21)	(1.60)	(1.63)
Compensation	−0.14	−0.16	−0.01	−0.03
	(0.11)	(0.11)	(0.14)	(0.14)
International	3.24*	3.35*	−1.41	−1.18
	(1.95)	(1.76)	(2.06)	(2.03)
Scpolitical	0.40	−0.02	0.17	−0.02
	(1.26)	(1.32)	(1.66)	(1.68)
Scother	1.80	2.04	2.60	2.36
	(8.01)	(7.22)	(1.87)	(1.88)
Scfirm	0.29	0.10	1.15	1.07
	(0.80)	(0.80)	(0.96)	(0.95)
Family	2.99	2.88	5.06**	5.33**
	(2.03)	(1.91)	(2.37)	(2.35)
Founder	9.72***	10.82***	3.02	2.28
	(3.51)	(3.25)	(2.23)	(2.36)
Firm age	−0.04	−0.03	0.32**	0.31**
	(0.07)	(0.06)	(0.16)	(0.17)
Size	1.44***	1.47***	0.75	0.74
	(0.53)	(0.52)	(0.80)	(0.78)
Ownership	1.35	1.24	2.86	3.04
	(2.27)	(2.11)	(2.41)	(2.41)
Top_holder	−1.45	−1.14	8.22	7.29
	(3.16)	(3.19)	(5.89)	(5.89)
Insti_holder	2.84	2.29	1.38	1.08
	(1.94)	(1.83)	(2.38)	(2.74)
Board size	1.02***	1.03***	1.08**	1.00**
	(0.33)	(0.33)	(0.46)	(0.45)
Independence	−4.34	−1.91	16.61	15.08
	(5.56)	(5.69)	(11.09)	(10.99)
Diversity	1.59	1.17	−6.03*	−5.82*
	(2.02)	(2.19)	(3.38)	(3.25)
ROA	−19.20**	−18.45**	12.73	12.84
	(7.57)	(7.44)	(10.28)	(10.25)
Leverage	0.14	0.70	−1.64	−2.09
	(3.35)	(3.33)	(5.44)	(5.33)
Revenue growth	2.35**	2.66***	2.31*	2.13*
	(0.92)	(0.95)	(1.23)	(1.22)
Current ratio	0.19	0.23	−0.72	−0.83*
	(0.29)	(0.27)	(0.48)	(0.47)
Pre_GDP	−6.50	−7.21	−0.61	0.60
	(4.76)	(4.54)	(9.38)	(9.36)
HHI	8.36	4.04	−15.42	−13.21
	(16.25)	(15.84)	(15.73)	(16.15)
Pre_CSR	0.17	0.17	0.31	0.33
	(0.12)	(0.12)	(0.30)	(0.30)
Turnover1 * Post1		2.84***
		(0.91)
Turnover2 * Post2				−1.86*
				(1.05)
Year fixed	Y	Y	Y	Y
Industry fixed	Y	Y	Y	Y
Province fixed	Y	Y	Y	Y
Constant	60.32	65.70	−19.43	−30.76
	(54.542)	(52.21)	(105.53)	(105.40)
Observations	705	705	618	618
R-squared	0.6700	0.6796	0.6191	0.6221

Note: This table reports the estimates of difference-in-differences (DID) regressions. Our first set of tests are based on a change of CEO from a nonhometown CEO to a hometown CEO. In this case, the control group consists of the firms that replaced a nonhometown CEO with another nonhometown CEO. The results of this set of tests are reported in columns 1 to 2. In the second set of tests, the treatment group consists of firms that replaced a hometown CEO with a nonhometown CEO, while the control group consists of firms that replaced a hometown CEO with another hometown CEO. The results of the second set of tests are reported in columns 3 to 4.

*p < .10. **p < .05. ***p < .01 (two-tailed tests).

ITCV

Following Frank (2000), Larcker and Rusticus (2010), and Badertscher, Katz, and Rego (2013), we also estimate the ITCV. The purpose of calculating this threshold is to assess the sensitivity of the results to the inclusion of potentially confounding variables (e.g., confounding omitted variables). In other words, ITCV suggests how large the endogeneity problem must be to overturn the results (e.g., making a significant coefficient insignificant). More specifically, ITCV is the product of two correlations: those of confounding variables with the dependent variable and with the independent variables (Frank, 2000).

The ITCV for CEO hometown identity in the CSR regression is 0.0657, which suggests that correlations of the unobserved confounding variable with CEO hometown identity and with CSR must be about 0.256 (i.e., $\sqrt{0.0657}$ ) to overturn the results reported in Table 2. The impact is significantly larger than that of Board size, the variable with the greatest influence on the CEO hometown identity coefficient in the CSR regression (impact = 0.0065).

Moreover, we also assessed the susceptibility of the CEO hometown identity effect to an omitted variable bias. Following Quigley, Hubbard, Ward, and Graffin (2020), we calculated the threshold for the percentage of bias required to invalidate the inference of within-firm effects of CEO hometown identity in the hybrid model. We carried out this test using the KonFound-it! statistical package (Frank, 2014). The result shows that, to invalidate our findings, 66.58% of observations would have to be replaced with observations for which the effect of CEO hometown identity on CSR was zero. Thus, it is unlikely that endogeneity issues could overturn our main findings.

Alternative measures of CSR

We employ three alternative measures of CSR. First, our theory suggests that hometown identity motivates the CEO to fulfill the expectations of fellow local townspersons. This implies at least a significant proportion of CSR investment targets local stakeholders. Therefore, we measure the level of local CSR by Local_economy, which captures the extent to which the focal firm has contributed to the local economy by involving multiple local stakeholders, including local employees, local suppliers/distributors, and the local community. We use the China Research Data Services Platform (CNRDS) to code a dummy variable of Local_economy as 1 if the firm has engaged in activities such as local employment, local procurement, and local distribution that enhance the local economy, and 0 otherwise.

Moreover, we used the natural logarithm of a firm's investment in environmental issues (environmental protection; pollution control) as another alternative measure of CSR. We obtained information about a firm's environmental investment from annual reports, CSR reports, environmental reports, and sustainability reports issued by firms. The results with this alternative measure of CSR are consistent with the results in the main analysis.

Finally, we obtain a comprehensive CSR score from Hexun (http://www.hexun.com). Hexun publishes professional evaluations of listed firms’ CSR in five dimensions: (1) responsibility to shareholders, (2) legal obligations to consumers and suppliers, (3) employee relationships, (4) environmental issues, and (5) taxation and donations. Each dimension includes detailed subindicators. As Hexun started issuing CSR for Chinese publicly listed firms from 2010, the time window for this supplementary analysis is 2010–2016. We use these data to create CSR measures: the total Hexun CSR (HCSRT, based on all the five dimensions). We chose this indicator because these activities target local stakeholders. For example, most of the directors, executives, and employees of Chinese listed companies are local (Knyazeva, Knyazeva, & Masulis, 2013; Ye, 2014).

Results with these alternative CSR measures are consistent with those in the main analyses.

Alternative measure of CEO hometown at the city level

Finally, as our measurement of CEO hometown identity is at the provincial level, we refined this measure by conducting an exhaustive search and obtaining hometown information at the municipal level for a subsample of 1,760 CEOs (45.8% of the total sample), with 869 CEOs steering firms in their hometown. We replicated the main results with this alternative measure.

Alternative estimation models

We also use OLS regression as an alternative way to test the relationship between CEO hometown identity and CSR, with cluster-adjusted firm-level and heteroskedasticity-robust standard errors. We obtain consistent results.

Discussion

This study examines whether and why CEO hometown identity can affect CSR. We predict that CSR initiatives can be motivated by a CEO's altruistic and long-term orientations caused by positive sentiments toward the CEO's hometown. We further explored the moderating effects of the firm's local connection and regional diversity of dialect. The empirical results suggest that CEOs’ hometown identity has a positive effect on CSR; this effect is stronger for firms that are more locally connected.

Theoretical Implications

Our study has important implications for upper echelons theory, social identity theory, and CSR research. First, we offer insights for upper echelons theory by examining the effect of CEOs’ hometown identity. This stream of research has examined either demographic factors or psychological factors (Chin et al., 2013; Petrenko et al., 2016; Tang et al., 2015, 2018). CEOs’ hometown identity has been largely neglected (Liu, Fisher, & Chen, 2018). Where CEOs were born and raised can have a significant influence on their reactions and consequently strategic decision-making. While psychological factors such as hubris and narcissism tend to reflect a CEO's self-serving motive, hometown identity triggers the CEO's altruistic and prosocial motive. Our investigation of hometown identity thus encourages students of upper echelons theory to explore the strategic implications of this unique factor further. Moreover, this study adds to social identity theory. While hometown has been treated in the literature as an extension of place identity (Makadok, Burton, & Barney, 2018), we advance this line of exploration by proposing a new construct of hometown identity and linking it to firms’ strategic decisions on CSR.

In addition, prior studies have examined the external and internal drivers of CSR. Internal factors include executive incentives (Deckop et al., 2006), CEOs’ political ideologies (Chin et al., 2013), slack resources (Waddock & Graves, 1997), and CEOs’ characteristics such as gender, education, personality, and experience, which have significant effects on CSR (Manner, 2010; Petrenko et al., 2016; Slater & Dixon-Fowler, 2009). We have investigated the effect of CEOs’ hometown identity on CSR. Hometown CEOs can be oriented to the long-term. Compared with nonhometown CEOs, they pay attention to the sustainable development of their hometowns and are less likely to harm the long-term interests of their hometowns in pursuit of short-term gains.

Further, we have identified the two moderators: firms’ local connection and regional dialect diversity. This resonates with the recent call for more investigations into the interaction between CEOs’ characteristics and contextual factors (Liu et al., 2018; Wowak et al., 2017). We find that links with local businesses strengthen the effect of CEO hometown identity on CSR, while effect of the hometown identity becomes weaker for firms located in a region with more diverse dialects. By identifying these boundary conditions, we provide a better understanding of the strategic implications of CEOs’ hometown identity for CSR.

Managerial Implications

Our findings have managerial implications for firms’ decisions when appointing CEOs and choosing the location of headquarters. First, we have shown that CEOs will promote CSR more actively when their firms are headquartered in their hometowns. In this regard, when appointing a CEO or choosing a location for their headquarters, firms should carefully consider their strategic priorities in the social domain. For example, a firm that puts a strategic emphasis on its CSR should probably hire a CEO whose hometown is where the firm is presently headquartered.

Second, our moderating analysis shows that when a firm has more local connections and is facing a more dialectally homogenous community, the positive effect of CEO hometown identity on CSR will be strengthened. Therefore, for those firms that want to improve their CSR, the key decision-makers should devote more attention to develop connections with the local stakeholders. Similarly, for any new CEO recruitment, when the candidate is a local person, the firm should also consider whether the region has more homogenous or heterogenous dialects.

Limitations and Future Research

Some limitations in this study may pave the road for future research. First, this study focuses on the relationship between CEOs’ hometown identity and CSR. A CEO's hometown identity can exert an influence on a wide spectrum of corporate strategic decisions (Hambrick & Mason, 1984). It would be fruitful for future research to explore whether, when, and why CEO hometown identity may affect other decisions of the firm and outcomes such as risk-taking and environmental performance.

Second, we have proposed that the relationship between CEO hometown identity and CSR is driven by the CEO's altruistic and prosocial motives resulting from a strong sentimental attachment to the hometown. Although our supplementary analysis with a survey data may have assessed this mechanism to some extent, we acknowledge some methodological limitations (such as samples and an endogeneity issue) associated with this research design. We hope that future research can reconfirm the proposed mechanism as well as its boundary conditions with more rigorous methods such as natural or laboratory experiments.

Third, the existing research has suggested some CEO-level psychological factors, such as hubris and narcissism (Tang et al., 2018), may also affect CSR. Then, it is naturally expected that these factors may play a role in how one associates with their hometown identity, especially if they see themselves as a prototypical or “elite” in-group member. However, due to the focus of this study as well as the data availability, we have not tested these ideas. Future research should take this route to explore all these possibilities.

Moreover, our measurements should be refined by future research. The main analysis has only tracked a CEO's hometown identity at the provincial level. We have adopted a more finely tuned CEO hometown identity measure at the municipal level for a subsample of CEOs in a robustness check and obtained consistent results. Future research should consider replacing our findings with a large-scale sample using refined measures. In addition, some CEOs may have moved with their parents from their hometown to other regions at an early age. These CEOs may identify less with the place where they were born but more with the place where they grew up and/or were educated. These possibilities should be tested by future research, and the operationalization of CEO hometown identity should certainly be refined.

On a final note, it would be naïve to assume that our results hold universally across different settings. Hometown identity may be stronger in China, as Chinese culture prioritizes collectivism and worship to the ancestor and past is deeply embedded in every single Chinese mind. Hometown therefore plays an especially important role in China. For example, for every Chinese lunar new year, most of Chinese (no matter where they are) will hurry back to their hometowns to celebrate the festivals with their relatives and fellow townspeople. Certainly, given its cultural uniqueness, hometown identity may play different roles across different cultural and institutional settings. We thus call for future research to confirm this conjecture.

Footnotes

Acknowledgments

We sincerely thank our Action Editor, Professor Aaron Hill, and the two anonymous reviewers for their guidance and support throughout the review process. We are also grateful for feedback provided by Lei Du, Yucai Hu, and Yue Wang. This research was supported by the National Science Foundation of China (Grant No. 71974205).

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

ORCID iDs

Helin Sun

Yi Tang

Notes

Appendix 1

Measurements in the Field Survey and Reliability/Validity Assessment

Measurements	SFL	Reference
Hometown Identity (Hometown): Cronbach's Alpha = .911, CR = .912, AVE = .563
1. When someone criticizes the hometown, it feels like a personal insult.	0.783	Mael and Ashforth (1992); Boivie et al. (2011); Lange, Boivie, and Westphal (2015); Farooq, Rupp, and Fraooq (2017)
2. I am very interested in what people think about the hometown.	0.773
3. When I talk about the hometown, I often say “we” rather than “they”.	0.765
4. When someone makes positive remarks about the hometown, it feels like a personal compliment.	0.782
5. If a story in the media criticized hometown, I would feel embarrassed, whether or not I know the person.	0.707
6. Being a member of the hometown is a major part of who I am.	0.699
7. Please indicate to what degree your self-image overlaps with the hometown's. (7-point response format)	0.741
8. Imagine that one of the circles on the left represents your self-definition or identity and the other circle on the right represents the hometown. Please indicate which case (A, B, C, D, E, F, G, or H) best describes the level of overlap between your self-definition and the hometown.	0.750
Altruistic Orientation (AO): Cronbach's Alpha = .853, CR = .852, AVE = .536		Kirchler (1997)
1. It is important to me to be able to make a contribution to hometown.	0.741
2. I will do my best to provide the needed help to my hometown.	0.774
3. I am ready to pay more money to hometown if necessary	0.700
4. In the eyes of my fellow villagers, they think I am a dedicated person.	0.671
5. Some behaviors of the company may hurt the hometown, but I think it is inevitable.	0.771
Long-term Orientation (LTO): Cronbach's Alpha = .835, CR = .837, AVE = .564
1. As your firm defines strategies, you generally emphasize long-term (over 5 years) goals and strategies.	0.806	Miller and Friesen (1982), Venkatraman (1989), Wang and Bansal (2012)
2. Your firm's criteria for resource allocation largely reflect long-term considerations.	0.781
3. Your firm emphasizes basic research to build future competitive edge.	0.679
4. As your firm defines strategies, your major concern is how to get short-term benefits(R).	0.731
Local: Cronbach's Alpha = 0.838, CR = .839, AVE = .565
1. Among our company's customers, there are more local customers.	0.742	Lane, Salk, and Lyles (2001); Stafford (1994)
2. Among our company's suppliers, there are more local suppliers.	0.776
3. Among our company's technical partners, there are more local partners.	0.727
4. Among our company's partners, there are more local companies.	0.761
Regional Dialect Diversity (Dialect): Cronbach's Alpha = 0.882, CR = .882, AVE = .651
1. There are many Chinese dialects in the area where our company is located.	0.827	Trax, Brunow, and Suedekum (2015)
2. The Chinese dialects in the area where our company is located are quite different.	0.781
3. The area where our company is located has a large number of migrants.	0.870
4. There is a big cultural difference between the foreigners in the area where our company is located and the local aboriginals.	0.745
CSR: Cronbach's Alpha = 0.892, CR = .851, AVE = .539
1. Our company participates in the activities which aim to protect and improve the quality of the local natural environment.	0.728	Turker (2009); Farooq, Payaud, Merunka, and Valette-Florence (2014); Farooq et al. (2017)
2. Our company implements special programs to minimize its negative impact on the local natural environment.	0.762
3. Our company contributes to the campaigns and projects that promote the well-being of the local community.	0.738
4. Our company's management goal is to reduce management costs rather than to pay attention to the needs and wishes of employees. (R)	0.738
5. Our company implements flexible policies to provide a good work and life balance for its local employees.	0.744
6. Our company protects local consumer rights beyond the legal requirements.	0.759
7. Our company provides full and accurate information about its products to its customers.	0.667
Model fit: χ²(444) = 721.3, p < .001; χ²/df = 1.62, GFI = .900, CFI = .960, IFI = .960, TLI = .955, NFI = .902, RMSEA = .039
Control variables
Gender1 = male，0 = female
CEO age1 = under 25 years, 2 = 26–35 years, 3 = 36–45 years, 4 = 46 years and above
Education1 = College degree and below, 2 = Undergraduate, 3 = Master, 4 = Doctorate
Founder1 = CEO is the founder of the firm, 0 = No
Firm age1 = 1–5 years, 2 = 5–10 years, 3 = 10–25 years, 4 = 25 years or more
Size1 = 100 people and below, 2 = 101–500 people, 3 = 501–1,000 people, 4 = 1,001 people or more
Industry1 = Agriculture, forestry, animal husbandry and fishery, 2 = Mining industry, 3 = Manufacturing industry, 4 = Electricity, heat, gas and water production and supply industry, 5 = Construction industry, 6 = Other industries
Ownership1 = State-owned enterprise, 2 = Private enterprise, 3 = Collective enterprise, 4 = Foreign-funded enterprise
Headquarters1 = The firm headquarters is in the CEO's hometown, 0 = No
Family1 = The firm is a family firm, 0 = No

Note: R, reverse coded item. All items are rated on a 7-point Likert scale (1 = “strongly disagree” and 7 = “strongly agree”), unless specified otherwise.

Appendix 2

Descriptive Statistics and Correlations of Variables in Field Survey

	(1)	(2)	(3)	(4)	(5)	(6)	(7)	(8)	(9)	(10)	(11)	(12)	(13)	(14)	(15)	(16)	(17)
(1) CSR	.73
(2) Hometown	.43	.75
(3) AO	.58	.62	.73
(4) LTO	.41	.40	.49	.75
(5) Local	.33	.34	.34	.24	.80
(6) Dialect	.05	.03	.03	.22	.04	.75
(7) Gender	.03	−.02	−.03	−.01	.01	.02
(8) CEO age	.02	.05	−.01	−.05	−.03	.04	.14
(9) Education	.04	.00	.00	.12	−.06	.04	.04	.10
(10) Founder	.20	.13	.11	.10	.07	.04	.08	−.09	−.02
(11) Size	.03	.01	.02	.09	−.03	.15	.08	.36	.12	−.20
(12) Firm age	−.03	.04	.04	.16	.05	.15	.09	.33	.03	−.30	.51
(13) Industry	−.16	−.11	−.16	−.26	−.02	−.20	−.11	−.04	−.17	−.08	−.21	−.15
(14) Region	.02	.06	.12	.07	.12	.06	−.04	−.14	−.06	−.09	.05	.03	.06
(15) Headquarters	.18	.12	.11	.18	.13	.12	.06	−.08	.06	.19	−.08	−.05	−.16	.07
(16) Ownership	.00	−.07	−.08	−.13	−.06	−.10	.01	.05	.03	.07	−.18	−.15	.09	−.08	−.10
(17) Family	.12	.11	.03	.06	.10	.10	.05	−.06	−.01	.22	.01	−.05	−.12	.01	.17	−.03
M	5.80	6.00	6.13	5.73	5.54	3.68	.58	2.60	2.17	.63	2.11	2.30	3.91	2.15	.76	2.01	.27
SD	.58	.56	.62	.79	.81	1.29	.49	.67	.63	.48	.80	.81	1.35	1.11	.43	.50	.44

Note: Values in bold on the diagonal represent the square root of AVE. Coefficients with a magnitude greater than 0.10 are significant at the .05 level.

References

Adams

Hardwick

1998. An analysis of corporate donations: United Kingdom evidence. Journal of Management Studies, 35(5): 641-654.

2013. Best-practice recommendations for defining, identifying, and handling outliers. Organizational Research Methods, 16(2): 270-301.

Altman

Low

S. M.

1992. Place attachment. New York: Plenum Press.

Antonsich

2010. Searching for belonging–an analytical framework. Geography Compass, 4(6): 644-659.

2008. Identification in organizations: An examination of four fundamental questions. Journal of Management, 34(3): 325-374.

Ashforth

B. E.

Mael

1989. Social identity theory and the organization. Academy of Management Review, 14(1): 20-39.

2013. The separation of ownership and control and corporate tax avoidance. Journal of Accounting and Economic, 56(2–3): 228-250.

Bammens

Hünermund

2020. Nonfinancial considerations in eco-innovation decisions: The role of family ownership and reputation concerns. Journal of Product Innovation Management, 37(5): 431-453.

Benabou

Tirole

2010. Individual and corporate social responsibility. Economica, 77(305): 1-19.

10.

Blau

P. M.

1977. Inequality and heterogeneity. Glencoe, IL: Free Press.

2011. Me or we: The effects of CEO organizational identification on agency costs. Academy of Management Journal, 54(3): 551-576.

2012. What motivates corporate managers to make socially responsible investments. SSRN Electron, 26: 164-181.

13.

Carroll

A. B.

1979. A three-dimensional conceptual model of corporate performance. Academy of Management Review, 4(4): 497-505.

2016. Sample selection bias and Heckman models in strategic management research. Strategic Management Journal, 37(13): 2639-2657.

2019. CEO tenure and corporate social responsibility performance. Journal of Business Research, 95: 292-302.

2013. Political ideologies of CEOs: The influence of executives’ values on corporate social responsibility. Administrative Science Quarterly, 58(2): 197-232.

17.

Chinese Academy of Social Sciences, & Australian Academy of Humanities. 1987. Atlas of Chinese language [M]. Hong Kong: Hong Kong Longman Publishing (Far East) Co., Ltd.

18.

Cobb

1977. The ecology of imagination in childhood. London: Routledge and Keegan Paul.

2011. Do powerful politicians cause corporate downsizing? Journal of Political Economy, 119(6): 1015-1060.

20.

Cooper

M. C.

1992. Environmental memories. In Altman

Low

(Eds.), Place attachment: 87-112. New York: Plenum Press.

21.

Crilly

Jiang

2016. Do-no-harm versus do-good social responsibility: Attributional thinking and the liability of foreignness. Strategic Management Journal, 37(7): 1316-1329.

2006. The effects of CEO pay structure on corporate social performance. Journal of Management, 32(3): 329-342.

1999. Qiaoxiang ties: Interdisciplinary approaches to “cultural capitalism” in South China. London and New York: Kegan Paul International.

2002. Beauty is in the eye of the beholder: The impact of organizational identification, identity, and image on the cooperative behaviors of physicians. Administrative Science Quarterly, 47(3): 507-533.

2014. Do better-connected CEOs innovate more? Journal of Financial and Quantitative Analysis, 49(5–6): 1201-1225.

2014. The impact of corporate social responsibility on organizational commitment: Exploring multiple mediation mechanisms. Journal of Business Ethics, 125(4): 563-580.

2017. The multiple pathways through which internal and external corporate social responsibility influence organizational identification and multifoci outcomes: The moderating role of cultural and social orientations. Academy of Management Journal, 60(3): 954-985.

2018. Social ties and favoritism in Chinese science. Journal of Political Economy, 126(3): 1134-1171.

2012. Psyched up or psyched out? The influence of coactor status on individual performance. Organization Science, 23(2): 402-415.

30.

Frank

K. A.

2000. Impact of a confounding variable on a regression coefficient. Sociological Methods Research, 29(2): 147-194.

31.

Frank

K. A.

2014. KonFound-it! Available at https://www.msu.edu/∼kenfrank/research.htm.

32.

Frijns

Dodd

Cimerova

2016. The impact of cultural diversity in corporate boards on firm performance. Journal of Corporate Finance, 41: 521-541.

33.

Galaskiewicz

1997. An urban grants economy revisited: Corporate charitable contributions in the twin cities, 1979–81, 1987–89. Administrative Science Quarterly, 42(3): 445-471.

34.

Garmache

D. L.

Neville

Bundy

Short

C. E.

2020. Serving differently: CEO regulatory focus and firm stakeholder strategy. Strategic Management Journal, 41(7): 1305-1335.

35.

Gropper

D. M.

Jahera

J. S.

Park

J. C.

2013. Does it help to have friends in high places? Bank stock performance and congressional committee chairmanships. Journal of Banking & Finance, 37(6): 1986-1999.

36.

Guthrie

2003. Survey on corporate-community relations. New York: Social Sciences Research Council.

37.

Hamamura

2017. Social identity and attitudes toward cultural diversity: A cultural psychological analysis. Journal of Cross-Cultural Psychology, 48(2): 184-194.

38.

Hambrick

Mason

1984. Upper echelons: The organization as a reflection of its top managers. Academy of Management Review, 9(2): 193-206.

39.

Hamel

Prahalad

C. K.

1989. Strategic intent. Harvard Business Review, 67(October): 63-76.

40.

Hamel

Prahalad

C. K.

1994. Competing for the future. Boston, MA: Harvard Business School Press.

41.

Heckman

J. J.

1979. Sample selection bias as specification error. Econometrica, 47(1): 153-161.

42.

Hernández

Martín

A. M.

Ruiz

Hidalgo

M. C.

2010. The role of place identity and place attachment in breaking environmental protection laws. Journal of Environmental Psychology, 30(3): 281-288.

43.

Hester

O’ Donnell

1987. Inner and outer landscapes: Integrating psychotherapy with design. Mass, 5: 30-36.

44.

Hill

A. D.

Johnson

S. G.

Greco

L. M.

O'Boyle

E. H.

Walter

S. L.

2020. Endogeneity: A review and agenda for the methodology-practice divide affecting micro and macro research. Journal of Management, 47(1): 105-143.

45.

Hillman

A. J.

Keim

G. D.

2001. Shareholder value, stakeholder management, and social issues: What’s the bottom line? Strategic Management Journal, 22(2): 125-139.

46.

Hodler

Raschky

P. A.

2014. Regional favoritism. Quarterly Journal of Economics, 129(2): 995-1033.

47.

Hogg

M. A.

Terry

D. J.

2000. Social identity and self-categorization processes in organizational contexts. Academy of Management Review, 25(1): 121-140.

48.

Hogg

M. A.

Terry

D. J.

White

K. M.

1995. A tale of two theories: A critical comparison of identity theory with social identity theory. Social Psychology Quarterly, 58(4): 255-269.

49.

Huang

Kisgen

D. J.

2013. Gender and corporate finance: Are male executives overconfident relative to female executives? Journal of Financial Economics, 108(3): 822-839.

50.

Huy

Q. N.

2011. How middle managers’ group-focus emotions and social identities influence strategy implementation. Strategic Management Journal, 32(13): 1387-1410.

51.

Jorgensen

B. S.

Stedman

R. C.

2001. Sense of place as an attitude: Lakeshore owners attitudes toward their properties. Journal of Environmental Psychology, 21(3): 233-248.

52.

Kang

2010. Stock-based CEO compensation and corporate attention to social responsibility. Working paper, Nanyang Technological University, Singapore.

53.

Kirchler

1997. The burden of new taxes: Acceptance of taxes as a function of affectedness and egoistic versus altruistic orientation. Journal of Socio-Economics, 26(4): 421-437.

54.

Knight

2006. Estimating the value of proposal power. American Economic Review, 95(5): 1639-1652.

55.

Knyazeva

Masulis

R. W.

2013. The supply of corporate directors and board independence. The Review of Financial Studies, 26(6): 1561-1605.

56.

Lai

Yang

Y. G.

2020. East, west, home's best: Do local CEOs behave less myopically? The Accounting Review, 95(2): 227-255.

57.

Lane

P. J.

Salk

J. E.

Lyles

M. A.

2001. Absorptive capacity, learning, and performance in international joint ventures. Strategic Management Journal, 22(12): 1139-1161.

58.

Lange

Boivie

Westphal

J. D.

2015. Predicting organizational identification at the CEO level. Strategic Management Journal, 36(8): 1224-1244.

59.

Larcker

D. F.

Rusticus

T. O.

2010. On the use of instrumental variables in accounting research. Journal of Accounting and Economic, 49(3): 186-205.

60.

J. W.

2020. A dual-agency model of firm CSR in response to institutional pressure: Evidence from Chinese publicly listed firms. Academy of Management Journal, 63(6): 2004-2032.

61.

Liang

K. Y.

Zeger

S. L.

1986. Longitudinal data analysis using generalized linear models. Biometrika, 73(1): 13-22.

62.

Lipsitz

S. R.

Fitzmaurice

G. M.

Orav

E. J.

Laird

N. M.

1994. Performance of generalized estimating equations in practical situations. Biometrics, 50(1): 270-278.

63.

Liu

Fisher

Chen

2018. CEO attributes and firm performance: A sequential mediation process model. The Academy of Management Annals, 12(2): 789-816.

64.

Lok

Willmott

2019. Embedded agency in institutional theory: Problem or paradox? Academy of Management Review, 44(2): 470-473.

65.

Lorenzo-Molo

C. F.

Udani

Z. A. S.

2013. Bringing back the essence of the “S” and “R” to CSR: Understanding the limitations of the merchant trade and the white man’s burden. Journal of Business Ethics, 117(1): 123-136.

66.

Luo

Wang

Zhang

2017. Whose call to answer: Institutional complexity and firms’ CSR reporting. Academy of Management Journal, 60(1): 321-344.

67.

Mael

Ashforth

B. E.

1992. Alumni and their alma mater: A partial test of the reformulated model of organizational identification. Journal of Organizational Behavior, 13(2): 103-123.

68.

Makadok

Burton

Barney

2018. A practical guide for making theory contributions in strategic management. Strategic Management Journal, 39(6): 1530-1545.

69.

Manner

M. H.

2010. The impact of CEO characteristics on corporate social performance. Journal of Business Ethics, 93(1): 53-72.

70.

Margolis

Walsh

2003. Misery loves companies: Rethinking social initiatives by business. Administrative Science Quarterly, 48(2): 268-305.

71.

Marquis

Glynn

M. A.

Davis

G. F.

2007. Community isomorphism and corporate social action. The Academy of Management Review, 32(3): 925-945.

72.

Marquis

Qian

2014. Corporate social responsibility reporting in China: Symbol or substance? Organization Science, 25(1): 127-148.

73.

Mattingly

J. E.

Berman

S. L.

2006. Measurement of corporate social action: Discovering taxonomy in the Kinder Lydenburg Domini ratings data. Business & Society, 45(1): 20-46.

74.

Maury

2006. Family ownership and firm performance: Empirical evidence from Western European corporations. Journal of Corporate Finance, 12(2): 321-341.

75.

McElroy

K. M.

Siegfried

J. J.

1986. The community influence on corporate contributions. Public Finance Quarterly, 14(4): 394-414.

76.

Mcwilliams

Siegel

2001. Corporate social responsibility: A theory of the firm perspective. Academy of Management Review, 26(1): 117-127.

77.

Miller

Friesen

P. H.

1982. Innovation in conservative and entrepreneurial firms: Two models of strategic momentum. Strategic Management Journal, 3(1): 1-25.

78.

Moore

2000. Placing home in context. Journal of Environmental Psychology, 20(3): 207-217.

79.

Nielsen-Pincus

Hall

Force

J. E.

Wulfhorst

J. D.

2010. Sociodemographic effects on place bonding. Journal of Environmental Psychology, 30(4): 443-454.

80.

O'Brien

R. M.

2007. A caution regarding rules of thumb for variance inflation factors. Quality & Quantity, 41:(5) 673-690.

81.

Petrenko

O. V.

Aime

Ridge

Hill

2016. Corporate social responsibility or CEO narcissism? CSR motivations and organizational performance. Strategic Management Journal, 37(2): 262-279.

82.

Podsakoff

P. M.

MacKenzie

S. B.

Lee

J. Y.

Podsakoff

N. P.

2003. Common method biases in behavioral research: A critical review of the literature and recommended remedies. Journal of Applied Psychology, 88(5): 879-903.

83.

Preacher

K. J.

Hayes

A. F.

2008. Asymptotic and resampling strategies for assessing and comparing indirect effects in multiple mediator models. Behavior Research Methods, 40(3): 879-891.

84.

Qian

2013. Diversity versus tolerance: The social drivers of innovation and entrepreneurship in US cities. Urban Studies, 50(13): 2718-2735.

85.

Qian

Zhu

2014. Chinese urban migrants’ sense of place: Emotional attachment, identity formation, and place dependence in the city and community of Guangzhou. Asia Pacific Viewpoint, 55(1): 81-101.

86.

Quigley

Hubbard

Ward

Graffin

2020. Unintended consequences: Information release and CEO stock option grants. Academy of Management Journal, 63(1): 155-180.

87.

Ratter

Gee

2012. Heimat–A German concept of regional perception and identity as a basis for coastal management in the Wadden Sea. Ocean & Coastal Management, 68: 127-137.

88.

Scannell

Gifford

2010. Defining place attachment: A tripartite organizing framework. Journal of Environmental Psychology, 30(1): 1-10.

89.

Semadeni

Withers

M. C.

Certo

S. T.

2014. The perils of endogeneity and instrumental variables in strategy research: Understanding through simulations. Strategic Management Journal, 35(7): 1070-1079.

90.

Slater

Dixon-Fowler

2009. CEO internal assignment experience and corporate social performance. Journal of Business Ethics, 89(3): 473-489.

91.

Sobel

M. E.

1982. Asymptotic confidence intervals for indirect effects in structural equation models. Sociological Methodology, 13: 290-312.

92.

Stafford

E. R.

1994. Using co-operative strategies to make alliances work. Long Range Planning, 27(3): 64-74.

93.

Tajfel

1974. Social identity and intergroup behaviour. Social Science Information, 13(2): 65-93.

94.

Tajfel

Turner

J. C.

1979. An integrative theory of intergroup conflict. In Austin

W. G.

Worchel

(Eds.) The Social Psychology of Intergroup Relations: 33-47. Monterey, CA: Brooks/Cole.

95.

Tang

Mack

Chen

2018. The differential effects of CEO narcissism and hubris on corporate social responsibility. Strategic Management Journal, 39(5): 1370-1387.

96.

Tang

Qian

Chen

Shen

2015. How CEO hubris affects corporate social (ir)responsibility. Strategic Management Journal, 36(9): 1338-1357.

97.

Tidwell

M. V.

2005. A social identity model of prosocial behaviors within nonprofit organizations. Nonprofit Management & Leadership, 15(4): 449-467.

98.

Trax

Brunow

Suedekum

2015. Cultural diversity and plant-level productivity. Regional Science and Urban Economics, 53: 85-96.

99.

Turker

2009. How corporate social responsibility influences organizational commitment. Journal of Business Ethics, 89(2): 189-204.

100.

Turner

J. C.

1982. Towards a cognitive redefinition of the social group. In Tajfel

(Ed.), Social identity and intergroup relationships: 15-40. Cambridge, UK: Cambridge University Press.

101.

Turner

J. C.

1985. Social categorization and the self-concept: A social cognitive theory of group behavior. In Lawler

E. J.

(Ed.), Advances in group processes: 2. 77-122. Greenwich, CT: JAI Press.

102.

Turner

J. C.

Hogg

M. A.

Oakes

P. J.

Reicher

S. D.

Wetherell

M. S.

1987. Rediscovering the Social group: A self-categorization theory. Oxford and New York: Basil Blackwell.

103.

Useem

1988. Market and institutional factors in corporate contributions. California Management Review, 30(2): 77-88.

104.

Uzzell

Pol

Badenas

2002. Place identification, social cohesion, and environmental sustainability. Environment and Behavior, 34(1): 26-53.

105.

Vaske

J. J.

Kobrin

K. C.

2001. Place attachment and environmentally responsible behavior. Journal of Environmental Education, 32(4): 16-21.

106.

Venkatraman

1989. Strategic orientation of business enterprises: The construct, dimensionality, and measurement. Management Science, 35(8): 942-962.

107.

Waddock

S. A.

Graves

S. B.

1997. The corporate social performance-financial performance link. Strategic Management Journal, 18(4): 303-319.

108.

Wang

Bansal

2012. Social responsibility in new ventures: Profiting from a long-term orientation. Strategic Management Journal, 33(10): 1135-1153.

109.

Wang

Qian

2011. Corporate philanthropy and corporate financial performance: The roles of stakeholder response and political access. Academy of Management Journal, 54(6): 1159-1181.

110.

Wernicke, G., Sajko, M., & Boone, C. 2022. How much influence do CEOs have on company actions and outcomes? The example of corporate social responsibility.” Academy of Management Discoveries, 8(1), 36–55.

111.

White

1980. A heteroskedasticity-consistent covariance matrix estimator and a direct test for heteroskedasticity. Econometrica, 48(4): 817-838.

112.

Wood

D. J.

1991. Corporate social performance revisited. Academy of Management Review, 16(4): 691-718.

113.

Wowak

A. J.

Gomez-Mejia

L. R.

Steinbach

A. L.

2017. Inducements and motives at the top: A holistic perspective on the drivers of executive behavior. Academy of Management Annals, 11(2): 669-702.

114.

Wowak

A. J.

Mannor

M. J.

Arrfelt

McNamara

2016. Earthquake or glacier? How CEO charisma manifests in firm strategy over time. Strategic Management Journal, 37(3): 586-603.

115.

Wright

2011. Diversity and the imagined community: Immigrant diversity and conceptions of national identity. Political Psychology, 32(5): 837-862.

116.

Miyata

1999. A dictionary of Chinese dialects [M]. Beijing: Zhonghua Book Company.

117.

2014. Independent director cash compensation and earnings management. Journal of Accounting and Public Policy, 33(4): 391-400.