Introduction
Research in strategic management has seen a burgeoning interest in the domain of corporations’ efforts to cater to a spectrum of stakeholder interests (Carroll, Reference Carroll1979; Flammer, Reference Flammer2015; Hawn & Ioannou, Reference Hawn and Ioannou2016; McWilliams & Siegel, Reference McWilliams and Siegel2001; Porter & Kramer, Reference Porter and Kramer2002; Wang, Choi, & Li, Reference Wang, Choi and Li2008). Acknowledging the important role of stakeholders in shaping firm strategy and performance, it is crucial to understand what drives a firm’s participation in addressing stakeholders’ demands and the initiatives it employs in this regard. Within this, extensive upper echelon literature suggests that CEOs, as the key locus of corporate control and governance, play a decisive role in determining whether firms respond to or overlook stakeholder interests (Chin, Hambrick, & Trevino, Reference Chin, Hambrick and Trevino2013; Petrenko, Aime, Ridge, & Hill, Reference Petrenko, Aime, Ridge and Hill2016; Tang, Qian, Chen, & Shen, Reference Tang, Qian, Chen and Shen2015). As part of this effort, recent research points out the significant role of CEOs’ prior experiences on corporate stakeholder engagement – including poverty experience (Han, Chi, & Zhou, Reference Han, Chi and Zhou2022; Xu & Ma, Reference Xu and Ma2022), natural disaster experience (O’Sullivan, Zolotoy, & Fan, Reference O’Sullivan, Zolotoy and Fan2021), and ideological imprinting experience (Marquis & Qiao, Reference Marquis and Qiao2020).
Although scholars have gained considerable insights into why CEOs’ prior experiences affect corporate stakeholder engagement (e.g., Han et al., Reference Han, Chi and Zhou2022; O’Sullivan et al., Reference O’Sullivan, Zolotoy and Fan2021; Xu & Ma, Reference Xu and Ma2022), scant attention has been devoted to understanding how CEOs, who are influenced by their prior experiences and guided by heterogeneous perceptions, allocate resources across distinct types of social activities for effective stakeholder management. The prevailing body of research appears to implicitly assume that CEOs do not discern between different types of social programs nor engage in efficiency assessment. This omission is surprising given that CEOs are key decision-makers of considerable engagement in social initiatives, and that their prior experiences not only shape their cognitions and beliefs concerning altruistic intent but also influence their perception of resource allocation (Choi, Shin, & Kim, Reference Choi, Shin and Kim2023; De Bruijn & Antonides, Reference De Bruijn and Antonides2022; Griskevicius et al., Reference Griskevicius, Ackerman, Cantú, Delton, Robertson, Simpson, Thompson and Tybur2013).
Our study attempts to advance the understanding of the asymmetric effects of CEO early-life experiences on corporate poverty alleviation efforts by studying how CEOs’ early-life poverty trauma exposure shapes resource allocation across various types of poverty alleviation programs in China. This context is suitable for several reasons. First, poverty alleviation initiatives address a spectrum of needs – from immediate relief to long-term goals (Jiang, Kang, Liang, & Yang, Reference Jiang, Kang, Liang and Yang2023; Liu, Feng, Wang, & Qiu, Reference Liu, Feng, Wang and Qiu2020). Thus, within the context of poverty alleviation, we can explore how CEOs allocate resources differently, drawing on their personal experiences. To elucidate, CEOs with early-life poverty trauma exposure demonstrate strong empathy toward people in poverty. It implies the notion that CEOs with early-life poverty trauma exposure possess ‘a set of processes, by which people come to expect others to be similar to themselves’ (Robbins & Krueger, Reference Robbins and Krueger2005: 32). They also have a comprehensive understanding of the poverty situation – what it entails and what people living in poverty need, and as a result, are more likely to have intrinsic motivation (Argote & Miron-Spektor, Reference Argote and Miron-Spektor2011; Deci, Koestner, & Ryan, Reference Deci, Koestner and Ryan1999). Based on the above reasons, CEOs’ perspectives and choices of different poverty alleviation programs are likely to vary. Second, poverty alleviation holds significant importance in China and is considered a semi-coercive institutional pressure for Chinese firms. Specifically, the Chinese government launched a nationwide targeted poverty alleviation campaign in 2016. This campaign mobilized organizations across society, including listed firms, providing a unique setting to study public firms’ participation in poverty alleviation. Firms in China employ diverse strategies when thinking about which poverty alleviation programs to engage in (Chang, He, & Wang, Reference Chang, He and Wang2021). For instance, some firms emphasize more on donation and philanthropy, and others may consider more about capability building. Therefore, we believe this is a timely and suitable context to test our research question.
Specifically, following the logic of Porter and Kramer (Reference Porter and Kramer2006), we classify poverty alleviation into generic involvement in poverty alleviation and strategic involvement in poverty alleviation. Generic involvement indicates that a firm adopts a value-transfer method to engage in poverty alleviation initiatives, whereas strategic involvement manifests a firm’s considerable effort to adopt a value-creation method to create shared value for firms and their stakeholders, i.e., groups vulnerable to poverty issues. Furthermore, in our context, generic involvement mainly consists of transient financial support, akin to a temporary infusion of resources such as donations, grants, or direct financial aid aimed at addressing immediate needs. This type of involvement is often perceived as less efficient in resource utilization because it does not integrate closely with the business practices of the firm or create long-term solutions. In contrast, strategic involvement in poverty alleviation is characterized by a more efficient use of resources that involves a better integration of the firm’s business operations with the needs of disadvantaged communities. It aims to create long-term social impact and offers long-term and sustainable solutions to communities subject to poverty issues. By doing this, these communities can sustain themselves and improve their circumstances continuously. The notion of this classification reflects the philosophy that firms and societies are mutually dependent on each other and ‘both business decisions and social policies must follow the principle of shared value’ (Porter & Kramer, Reference Porter and Kramer2006: 82). It is also consistent with the practices in prior literature (e.g., Cuypers, Koh, & Wang, Reference Cuypers, Koh and Wang2016), suggesting that corporate philanthropy is associated with quantitative (the amount of giving) and qualitative (the type of giving) information cues, which can be perceived as substantive rather than symbolic.
We develop our arguments by considering the roles of both altruistic motives and emphasis on resource efficiency in shaping CEOs’ perceptions and appraisals of distinct poverty alleviation initiatives (Cannon, Goldsmith, & Roux, Reference Cannon, Goldsmith and Roux2019; Choi et al., Reference Choi, Shin and Kim2023; De Bruijn & Antonides, Reference De Bruijn and Antonides2022). We postulate that CEOs with greater early-life poverty trauma exposure exhibit heightened altruism and empathy, as well as a deeper understanding of poverty, which in turn leads to greater involvement in both generic and strategic poverty alleviation efforts. Moreover, CEOs with early-life poverty trauma exposure may place emphasis on resource efficiency when deliberating on such poverty alleviation commitments (De Bruijn & Antonides, Reference De Bruijn and Antonides2022). These CEOs are expected to devote increased efforts toward associating poverty alleviation involvement with value creation and emphasizing the resource efficiency of the involvement. Since strategic involvement in poverty alleviation would to a larger extent optimize the efficiency of firm resources, we propose that CEOs with early-life poverty trauma exposure would engage more in strategic initiatives than in generic ones. We provide a supplementary analysis of how emphasis on resource efficiency mediates the relationship between CEOs’ early-life poverty trauma exposure and the varying participation of firms in different types of poverty alleviation programs.
Our study makes several contributions to extant literature. First, our study extends the understanding of how CEOs’ early-life experiences influence firms’ social practice engagement by exploring the underlying psychological mindsets. Prior studies primarily adopt the prosocial orientation lens to investigate the influence of CEO early-life experiences on general social practice engagement. Our study offers new insights into strategic management literature by showing that CEOs’ early-life trauma experience may influence their mindfulness regarding resource allocation efficiency and further shape the extent to which they engage in social practice in a strategic way. Second, we extend the literature on the approach to stakeholder engagement. Previous studies primarily focus on distinguishing between symbolic and substantive social practices. Our research advances this stream of literature by exploring the heterogeneity of substantive social practices through the lens of comparison between generic and strategic involvement in poverty alleviation. Third, our study echoes the theme of grand challenges and the UN Sustainable Development Goals (SDGs) by contributing to the growing literature on social practice in transforming economies, particularly in China. Building on this context, we enhance the understanding of poverty alleviation’s multidimensional nature, its inherent complexity, and the context-relatedness between CEOs’ early-life experiences and their involvement in poverty alleviation efforts.
Theoretical Background and Hypotheses Development
CEOs’ Early-Life Experience
Grounded in the pioneering work by Hambrick and Mason (Reference Hambrick and Mason1984), the empirical research trajectory on the upper echelons of organizations has significantly advanced. This theoretical framework integrates with the notion of bounded rationality, suggesting that executives’ pursuit of rational decision-making is often constrained by informational deficits, time restrictions, cognitive boundaries, and situational complexity (Cyert & March, Reference Cyert and March1963). Hence, it has become widely recognized in strategic management discourse that executives, through their distinct characteristics, values, and cognitive patterns, interpret and navigate firm-specific challenges, inextricably influencing strategic direction and organizational performance. Research probing the features of CEOs has primarily bifurcated along two pathways (Finkelstein, Hambrick, & Cannella, Reference Finkelstein, Hambrick and Cannella2009). One pathway is largely centered on examining the strategic implications of psychological properties and personal values such as the Big Five personality dimensions (Herrmann & Nadkarni, Reference Herrmann and Nadkarni2014), overconfidence (Chen, Crossl, & Luo, Reference Chen, Crossl and Luo2015), and political ideology (Chin et al., Reference Chin, Hambrick and Trevino2013; Christensen, Dhaliwal, Boivie, & Graffin, Reference Christensen, Dhaliwal, Boivie and Graffin2015). Concurrently, an emerging strand of literature focuses on exploring the impact of prior experiences on executives’ strategic decision-making paradigms (Malmendier, Tate, & Yan, Reference Malmendier, Tate and Yan2011; Miller & Shamsie, Reference Miller and Shamsie2001; O’Sullivan et al., Reference O’Sullivan, Zolotoy and Fan2021).
More specifically, extant research has explored how behaviors and preferences are shaped by experiences such as career diversity (Crossland, Zyung, Hiller, & Hambrick, Reference Crossland, Zyung, Hiller and Hambrick2014), military service (Koch-Bayram & Wernicke, Reference Koch-Bayram and Wernicke2018), and educational backgrounds (Bai, Tsang, & Xia, Reference Bai, Tsang and Xia2020), among others. Among this stream of literature, scholars have further examined the impact of imprinting events that occurred during executives’ early life and were less influenced by individual motivations and choices (Campbell, Jeong, & Graffin, Reference Campbell, Jeong and Graffin2019). For instance, Kish-Gephart and Campbell (Reference Kish-Gephart and Campbell2015) empirically established that executives’ formative experiences with social class during their childhood exert a lingering and differentiated influence on their predilections. A background in an upper social class upbringing intensifies executives’ sense of security and control, consequently amplifying their propensity for risk-taking behaviors. In a parallel vein, Campbell et al. (Reference Campbell, Jeong and Graffin2019) unveiled the strategic ramifications of CEOs’ ordinal position in their birth orders, asserting that CEOs born earlier demonstrate a lower risk appetite compared to those born later. In a more related context, research investigating the impact of CEOs’ early-life experiences on CSR has revealed various correlations. For instance, CEOs who experienced famine during their childhood tend to make significant philanthropic contributions (Han et al., Reference Han, Chi and Zhou2022). Similarly, CEOs who grew up in poverty or have a lower childhood socioeconomic status are linked to enhanced environmental, social, and governance (ESG) performance (Kahhali, Reference Kahhali2022; Liu, Zhang, & Zhang, Reference Liu, Zhang and Zhang2024). Furthermore, Choi et al. (Reference Choi, Shin and Kim2023) found a positive association between CEOs’ childhood experiences of natural disasters and their firms’ CSR dedication. CEOs who were themselves ‘Sent-down youth’ have a positive impact on CSR engagement (Li, Jiang, Zhang, Huang, & Wang, Reference Li, Jiang, Zhang, Huang and Wang2023). In contrast, Fang, Khoo, Luo, Pathan, and Wang (Reference Fang, Khoo, Luo, Pathan and Wang2023) observed that firms led by CEOs who experienced famine tend to show poorer CSR performance.
Our exploration aligns with the imprinting theory framework, which asserts the profound influence of sensitive periods in an individual’s life. Imprinting is described as a process whereby ‘a focal entity develops characteristics that reflect prominent features of the environment, and these characteristics continue to persist despite significant environmental changes in subsequent periods’ (Marquis & Tilcsik, Reference Marquis and Tilcsik2013: 199). This theory has been applied extensively in management literature, sparking particular interest in the early-life experiences of executives. This research has shed light on how these formative experiences inform executive leadership styles, decision-making processes, and strategic orientations (Higgins, Reference Higgins2005; Holburn & Zelner, Reference Holburn and Zelner2010; McEvily, Jaffee, & Tortoriello, Reference McEvily, Jaffee and Tortoriello2012). Social psychology perspectives further illuminate childhood as a pivotal stage in sculpting individuals’ cognitive frameworks and values, with adversity faced during this period often leaving a profound, enduring impact into adulthood (Griskevicius et al., Reference Griskevicius, Ackerman, Cantú, Delton, Robertson, Simpson, Thompson and Tybur2013; Miller, Chen, & Parker, Reference Miller, Chen and Parker2011). Expanding upon this rich body of literature, our research uniquely focuses on the distinct influence of executives’ childhood experiences of poverty on corporate social strategies. Specifically, we scrutinize how these foundational experiences shape firms’ engagements in poverty alleviation initiatives, furthering our understanding of the long-lasting effects of early-life imprinting on strategic decision-making processes.
Poverty Alleviation in China
Poverty represents a universal challenge that confronts nations and the global community collectively (Haushofer & Fehr, Reference Haushofer and Fehr2014). Poverty alleviation directly addresses the first SDG (SDG 1: No Poverty) set by the United Nations. This alignment positions poverty alleviation efforts as a critical component of global sustainable development initiatives. The pursuit of development, the eradication of poverty, and the attainment of shared prosperity embody the enduring aspirations of humankind (Bapna, Reference Bapna2012). In 2013, President Xi Jinping introduced the concept of ‘targeted poverty alleviation’ as a key component of China’s comprehensive anti-poverty strategy and in alignment with the Communist Party of China’s centenary goal of building a ‘moderately prosperous society’.Footnote 1 To implement the poverty alleviation agenda, China emphasizes the importance of accurate poverty identification, appropriate project arrangements, and effective implementation to ensure that assistance and resources reach poverty-stricken villages and households efficiently (Jiang et al., Reference Jiang, Kang, Liang and Yang2023).
In 2016, the Chinese government launched a nationwide campaign targeting poverty alleviation, aiming to eradicate poverty by 2020. Recognizing the pivotal role of the financial capital market in this endeavor, listed firms were actively motivated to fulfill their social obligations (Chang et al., Reference Chang, He and Wang2021). In September 2016, the China Securities Regulatory Commission strongly advocated and incentivized listed firms to assume social responsibility in reducing poverty. Subsequently, in December 2016, the Shanghai and Shenzhen Stock Exchanges issued a Notice on Disclosing Firms’ Poverty Alleviation Information to encourage listed firms to actively engage in poverty alleviation initiatives. These guidelines stipulated that firms should disclose detailed information about their poverty alleviation endeavors in their annual reports or CSR reports, including quantitative information on their pecuniary and nonpecuniary contributions to the program and the outcomes achieved in lifting individuals out of poverty. The government’s registration and tracking of individuals living in poverty serve to enhance the credibility of firms’ disclosures, effectively mitigating the prevalent issue of boilerplate language in previous CSR reports.
The adoption of poverty alleviation programs by corporations is gaining popularity as an emerging manifestation of CSR in China (Zhang, Huang, Wu, Wang, Nygaard, & Qiu, Reference Zhang, Huang, Wu, Wang, Nygaard and Qiu2022). In light of the institutional pressure exerted by the poverty alleviation policy of the Chinese government, significant heterogeneity persists in the extent to which listed firms conform to this policy and engage in poverty alleviation activities. Scholars have directed their attention toward understanding the motivations that drive firms to participate in targeted poverty alleviation activities. Notably, previous studies have demonstrated that political pressure from the government serves as the primary driver behind firms’ spending on poverty alleviation initiatives (e.g., Huang, Song, Lv, & Zhou, Reference Huang, Song, Lv and Zhou2019). Furthermore, their study reveals that such spending increases with firms’ visibility and available resources.
Our study attempts to extend the emerging stream of research on how CEOs’ early-life experiences contribute to the formation of a firm’s prosocial behaviors (Han et al., Reference Han, Chi and Zhou2022; Kahhali, Reference Kahhali2022; Liu et al., Reference Liu, Zhang and Zhang2024; O’Sullivan et al., Reference O’Sullivan, Zolotoy and Fan2021). Previous studies rarely consider the multiplicity of corporate social activities. To this end, our study builds upon the multiplicity of poverty alleviation programs and proposes that CEOs’ early-life poverty trauma exposure may shape firms’ emphasis on generic versus strategic involvement in poverty alleviation.
CEOs’ Early-Life Poverty Trauma Exposure and Asymmetric Involvement in Poverty Alleviation
We propose that CEOs with greater early-life poverty trauma exposure are more inclined to engage in poverty alleviation initiatives due to their higher level of empathy and emotional connection to the impoverished (Cameron, Hutcherson, Ferguson, Scheffer, Hadjiandreou, & Inzlicht, Reference Cameron, Hutcherson, Ferguson, Scheffer, Hadjiandreou and Inzlicht2019; Cialdini, Schaller, Houlihan, & Arps, Reference Cialdini, Schaller, Houlihan and Arps1987). These feelings and cognitions are considered to be shaped by a transformative journey, heavily influenced by CEOs’ own experiences and observations of poverty during their formative years, and further promote prosocial behaviors (Stellar et al., Reference Stellar, Gordon, Piff, Cordaro, Anderson, Bai, Maruskin and Keltner2017). The exposure to poverty fosters a sense of relatedness and interconnectedness with those in similar situations (Stewart, Makwarimba, Reutter, Veenstra, Raphael, & Love, Reference Stewart, Makwarimba, Reutter, Veenstra, Raphael and Love2009). This perception can be instrumental in projecting their favoritism and commitment to poverty alleviation campaigns (Robbins & Krueger, Reference Robbins and Krueger2005). Therefore, CEOs with early-life poverty trauma exposure are more likely to channel their efforts toward poverty alleviation, making a significant impact on the lives of the less privileged.
In our context, we focus on two types of involvement in poverty alleviation: generic and strategic. Generic involvement in poverty alleviation refers to the temporary infusion of resources, such as donations, grants, or direct financial aid, aimed at addressing immediate needs. In contrast, strategic involvement in poverty alleviation aims to help poverty-stricken communities achieve self-sufficiency and sustainable development. As aforementioned, since CEOs with early-life poverty trauma exposure may exhibit greater empathy and emotional connections to the impoverished, we expect these CEOs to engage more in both generic and strategic poverty alleviation activities. Thus, we propose the following hypothesis:
Hypothesis 1 (H1): Ceteris paribus, CEOs’ early-life poverty trauma exposure is positively associated with generic involvement in poverty alleviation.
Hypothesis 2 (H2): Ceteris paribus, CEOs’ early-life poverty trauma exposure is positively associated with strategic involvement in poverty alleviation.
Thus far, we have established that CEOs with greater early-life poverty trauma exposure may engage more in both generic and strategic poverty alleviation initiatives. However, it is crucial to recognize that their emphasis on these initiatives may not be uniform. These CEOs may exhibit trade-off thinking when considering different types of poverty alleviation initiatives (De Bruijn & Antonides, Reference De Bruijn and Antonides2022; Huang et al., Reference Huang, Song, Lv and Zhou2019; Shah, Shafir, & Mullainathan, Reference Shah, Shafir and Mullainathan2015), i.e., the prioritization between generic and strategic involvement in poverty alleviation. We propose that CEOs with early-life poverty trauma exposure will exhibit a stronger inclination toward strategic involvement in poverty alleviation compared to generic involvement. This is based on the assertion that individuals with exposure to poverty trauma will place more emphasis on resource efficiency because poverty is usually associated with resource constraints, and these constraints force individuals to focus on promoting resource efficiency (De Bruijn & Antonides, Reference De Bruijn and Antonides2022; Krautkraemer, Reference Krautkraemer1998). Such an emphasis on resource efficiency results in individuals conducting careful and complex considerations of how resources should be consumed.
In the context of poverty alleviation, we suggest that CEOs with greater early-life poverty trauma exposure will prioritize strategic involvement over generic involvement. They may place greater emphasis on resource efficiency and make more deliberative and comprehensive evaluations of their participation in poverty alleviation initiatives. Generic involvement in poverty alleviation mainly provides immediate relief and creates dependency on such relief and is usually perceived as lacking efficient use of financial resources (Cuypers et al., Reference Cuypers, Koh and Wang2016; Kaul & Luo, Reference Kaul and Luo2018). In contrast, strategic involvement in poverty alleviation is a more efficient way of resource allocation because it has better integration between poverty communities’ needs and firms’ business operations and simultaneously creates a larger long-term social impact, promoting self-sufficiency among the economically disadvantaged (Cuypers et al., Reference Cuypers, Koh and Wang2016; Kaul & Luo, Reference Kaul and Luo2018; Vishwanathan, van Oosterhout, Heugens, Duran, & Van Essen, Reference Vishwanathan, van Oosterhout, Heugens, Duran and Van Essen2020).
Moreover, CEOs with early-life poverty trauma exposure possess a profound understanding of poverty and its intricate, deeply entrenched nature, leading them to favor strategic poverty alleviation initiatives that enable impoverished communities to take control of their economic development and futures as well as realize self-reliance (Xu & Ma, Reference Xu and Ma2022; Zahra, Neubaum, & Hayton, Reference Zahra, Neubaum and Hayton2020; Zhang et al., Reference Zhang, Wang and Jia2022). Relatedly, the culture of poverty theory suggests that impoverished communities develop certain attitudes, behaviors, and ways of thinking that make it harder for them to escape poverty (Harvey & Reed, Reference Harvey and Reed1996). CEOs with early-life exposure to poverty adversity care more about how resources can be utilized to help these communities achieve the status of self-sufficiency. In addition, such exposure informs a more nuanced understanding of which resources are truly essential, leading to more effective prioritization in resource allocation decisions (De Bruijn & Antonides, Reference De Bruijn and Antonides2022; Gupta, Briscoe, & Hambrick, Reference Gupta, Briscoe and Hambrick2018). These CEOs may also recognize that immediate relief measures alone cannot efficiently combat poverty. For instance, Zong Qinghou’s personal experiences with poverty have profoundly shaped his understanding of effective poverty alleviation strategies.Footnote 2 Rather than merely providing immediate relief such as donations, Zong emphasizes the necessity of enhancing the ‘hematopoietic’ capacity of local communities (i.e., self-sufficient ability). His approach to poverty alleviation predominantly focuses on industrial development, introducing advanced management approaches, and fostering local enterprises. Thus, we propose the following hypothesis:
Hypothesis 3 (H3): Ceteris paribus, the positive influence of CEOs’ early-life poverty trauma exposure on strategic involvement in poverty alleviation is stronger than that on generic involvement in poverty alleviation.
Methods
Data and Sample
Our initial sample included all publicly listed Chinese firms from 2016 to 2021. The choice of 2016 as the starting point for the observation window was predicated on the availability of data pertaining to firms’ poverty alleviation efforts. We extracted data on firms’ involvement in various poverty alleviation tactics from the China Stock Market and Accounting Research (CSMAR) database. We then obtained data about the CEO’s birth information, namely the birth year and birth province, from the CSMAR database to assemble an understanding of the CEOs’ early-life poverty trauma exposure. Data relating to CEO characteristics, firm-level variables, and province-level variables were garnered from the CSMAR database.
We constructed our sample through several steps. We first classified CEOs who were aged 7–11 years during the famine (i.e., 1959–1961) as the treatment group, which includes 172 firms and 173 CEOs during our sample period. We then classified CEOs who were born within three years after the famine (i.e., 1962–1964) as the control group, which includes 1339 firms and 1496 CEOs. We then identified the famine severity for each CEO based on their birthplace information. Our sample was reduced to 391 firms and 407 CEOs due to missing information on CEOs’ birthplaces. After merging the CEO-level data with firm-level and province-level data, we ended up with a sample of 1246 firm-year observations derived from 375 distinct firms. We winsorized all the continuous variables at the 1% level to minimize the potential influence of outliers (Jia, Ruan, & Zhang, Reference Jia, Ruan and Zhang2017; Xiang, Jia, & Zhang, Reference Xiang, Jia and Zhang2022).
Measures
Dependent variables
Generic poverty alleviation. Generic involvement in poverty alleviation predominantly entails transient financial interventions, such as donations, grants, or direct financial aid, which are designed to address immediate needs through temporary resource infusion. Social security poverty alleviation, healthcare poverty alleviation, and environmental poverty alleviation fall into this category. These programs often transfer resources directly to the poor, and such direct giving is more prone to being perceived as lacking deliberation, and inefficient in the allocation of considerable financial resources (Cuypers et al., Reference Cuypers, Koh and Wang2016). We summed up the amount of firms’ engagement in these three categories to measure Generic poverty alleviation and then standardized this variable.
Strategic poverty alleviation. Strategic poverty alleviation is distinguished by a more efficient allocation of resources through the integration of a firm’s business operations with the needs of disadvantaged communities. This approach focuses on creating long-term social impact by providing sustainable solutions that enable these communities to continuously improve their circumstances and sustain themselves over time. This strategic approach ensures that the interventions are not only impactful but also contribute to the sustained development of the communities involved. Industrial development poverty alleviation, education poverty alleviation, and vocational skill training poverty alleviation fall into this category. We summed up the amount of firms’ engagement in these three categories to measure Strategic poverty alleviation and then standardized this variable.
Industrial development, characterized by the establishment of plants and enterprises, emerges as a pivotal component of our strategic involvement category. It plays a crucial role in poverty alleviation (Huang et al., Reference Huang, Song, Lv and Zhou2019). By creating avenues for employment and self-employment, industrial development propels capital accumulation and regional economic growth. This dynamism not only results in a sustainable uplift in income within poverty-ridden regions but also triggers transformative changes in the social structures that perpetuate poverty (Kao, Chen, Wu, & Yang, Reference Kao, Chen, Wu and Yang2016). Furthermore, firms find value creation more efficient through this innovative avenue, thereby effectively leveraging their expertise and experiences (Huang et al., Reference Huang, Song, Lv and Zhou2019). Education is another vital component of strategic involvement in poverty alleviation. With its economic and social functions, education fosters skills and knowledge in individuals, thereby improving their employment prospects and earning potential. Notably, education holds the unique capacity to interrupt the intergenerational cycle of poverty (Li, Wang, Cheng, Shuai, Shuai, & Liu, Reference Li, Wang, Cheng, Shuai, Shuai and Liu2020), thus improving the efficiency of poverty alleviation efforts. Simultaneously, vocational skill training forms an essential facet of poverty alleviation, equipping the poor with the skills and knowledge required to address multidimensional challenges. Increased employment opportunities and wealth creation can effectively combat poverty (Saci, Reference Saci2023). We further offered explicate examples as a reflection of strategic poverty alleviation. For instance, in 2021, Huawei officially launched the ‘Tech School’ project, leveraging its innovative practices and expertise in the field of information and communication technology to develop a series of technology-focused courses. As reported, Ling Zhang, a professor at the School of Education at Ningxia University, also noted, ‘The Huawei Tech School allows our children to experience the power of technology through real-life scenarios. Planting such seeds during their primary school years can have a profound impact on their futures. While it may seem like a small classroom, it is driving a big dream in tech education’. Thus, such education and vocational skills training initiatives in poverty alleviation helped nurture local digital talent. These projects were directly related to the company’s core business and laid the groundwork for potential development in the future.Footnote 3 In the field of industrial poverty alleviation, Alibaba’s Rural Taobao project served as a compelling example.Footnote 4 As a strategic initiative of Alibaba Group, Rural Taobao collaborated extensively with local governments to establish a dual-level service network spanning counties and villages, leveraging the e-commerce platform to its full advantage.
Independent variable
CEO early-life poverty trauma exposure. We utilized the severity of a CEO’s early-life famine experience as an indicator of their exposure to early-life poverty trauma. This variable comprised two components: whether the CEO personally experienced the famine and the severity of the famine in the CEO’s birth province (Feng & Johansson, Reference Feng and Johansson2018; Han et al., Reference Han, Chi and Zhou2022; Zhang, Reference Zhang2017). First, we constructed a binary variable named CEO early-life poverty experience to capture whether the CEO experienced the famine. This variable equals 1 if CEOs were within the age group of 7–11 between the years 1959 and 1961 (born 1948–1954), and equals 0 if CEOs were born within 3 years post-1961.Footnote 5 We chose the age range of 7–11 for our treatment group of CEOs because this period, known as ‘middle childhood’ or the ‘entry into the age of reasoning’, is crucial for significant cognitive and brain development (Cheng, Guo, Hayward, Smyth, & Wang, Reference Cheng, Guo, Hayward, Smyth and Wang2021; Han et al., Reference Han, Chi and Zhou2022). This developmental stage is particularly sensitive to environmental inputs, with experiences during this time persisting in memory and shaping personality (DelGiudice, Reference DelGiudice2018; Roberts & DelVecchio, Reference Roberts and DelVecchio2000). Furthermore, traits established during this period tend to remain stable and persist into adulthood, influencing future actions and decisions (Gerrig & Zimbardo, Reference Gerrig and Zimbardo2010; Thompson & Nelson, Reference Thompson and Nelson2011). We treated CEOs born within three years after 1961 as the control group to select CEOs who did not experience the famine. In addition, this period could also exclude the effect of other significant social traumas in China, such as the Cultural Revolution that occurred in 1966. This operationalization allowed us to mitigate the bias possibly caused by confounding events (Bertrand, Duflo, & Mullainathan, Reference Bertrand, Duflo and Mullainathan2004).
Second, we quantified the severity of the famine in the CEOs’ birth provinces using the excess death rate of the province during the famine (Han et al., Reference Han, Chi and Zhou2022; Long, Tian, Hu, & Yao, Reference Long, Tian, Hu and Yao2020). We calculated it as famine severity = ([average death rate during the famine period (1959–1961) − average death rate for the three years prior to the famine (1956–1958)] + [average death rate during the famine period (1959–1961) − average death rate for the three years after the famine (1962–1964)])/2 (Han et al., Reference Han, Chi and Zhou2022: 1608). Figure 1 illustrates the magnitude of famine severity across provinces. Finally, we measured the independent variable, CEO early-life poverty trauma exposure, as the product of the values assigned to CEO early-life poverty experience and famine severity. By doing this, our measure can capture not only the age effect but also the severity difference across regions (Feng & Johansson, Reference Feng and Johansson2018; Long et al., Reference Long, Tian, Hu and Yao2020).
The magnitude of famine severity across provinces
Control Variables
As for firm-level characteristics, we first controlled for the influence of state ownership. State-owned enterprises (SOEs) may have a great effect on poverty alleviation, owing to the political pressure exerted by government entities (Chang et al., Reference Chang, He and Wang2021). Thus, we included an SOE variable, assigning it a value of 1 if the firm was ultimately controlled by the government or its agency institution and 0 otherwise. The financial standing of firms, which could impact their capacity for poverty alleviation, was also taken into account (Xiang et al., Reference Xiang, Jia and Zhang2022). We controlled for firm performance, which was operationalized as return on assets (ROA). Given that older firms have demonstrated more engagement in prosocial behavior and paid more attention to environmental protection than human rights (Withisuphakorn & Jiraporn, Reference Withisuphakorn and Jiraporn2016), we thus incorporated firm age, measured by the number of years since the firm’s establishment, as a control variable. Meanwhile, firms with different growth opportunities may adjust their resource allocation to poverty alleviation differently (Chang et al., Reference Chang, He and Wang2021). Thus, we controlled for firm growth opportunities, termed as revenue growth, which we quantified as the revenue growth rate. Recognizing that larger firms generally demonstrate increased engagement in poverty alleviation due to their enhanced visibility and extensive resources (e.g., Chang et al., Reference Chang, He and Wang2021; Xiang et al., Reference Xiang, Jia and Zhang2022), we thus included firm size, measured as the natural log-transformed total assets. Following previous studies (e.g., Xiang et al., Reference Xiang, Jia and Zhang2022; Xu & Ma, Reference Xu and Ma2022), we further included Cash flow, measured as corporate cash flow scaled by total assets, Selling expense ratio, measured as dividing selling expenses by operating revenue, and R&D expense ratio, measured as dividing research and development expenses by operating revenue. Recognizing that a firm’s ownership structure may influence its poverty alleviation decisions, we also included Manager share ratio and Largest shareholder shareholding ratio.
We also included some CEO-level control variables. CEOs’ education degrees would affect firms’ prosocial behavior (Xu & Ma, Reference Xu and Ma2022). We thus controlled for CEO education, assigning the value of 1 if the CEO held a bachelor’s degree or above and 0 otherwise. Another vital factor considered was the CEO’s gender. Empirical evidence suggested that female CEOs and male CEOs show different preferences for prosocial behavior (Han et al., Reference Han, Chi and Zhou2022; Tang et al., Reference Tang, Qian, Chen and Shen2015). Consequently, we incorporated CEO gender as a binary variable, coded as 1 for a female CEO and 0 otherwise. In addition, we included a binary variable for CEO duality, a situation wherein the CEO concurrently served as the chairman of the firm. This dual role could potentially augment the CEO’s effect on the firm’s prosocial activities due to the consolidation of decision-making power vested in the CEO (Han et al., Reference Han, Chi and Zhou2022; Xu & Ma, Reference Xu and Ma2022). We further included political connection as another control variable. Executives with higher levels of political connection would perceive higher pressure to engage in poverty alleviation from the Chinese central government. In China, a firm’s political connections are commonly manifested by whether the CEO or chairman has served or is currently serving as a delegate of the People’s Congress or the Chinese People’s Political Consultative Conference (CPPCC) at one of four hierarchical levels (i.e., from county to national level). We thus measured Political connection level with categorical variable, assigning values of 1, 2, 3, or 4 corresponding to each level, with a value of 0 indicating the absence of political connection. For corporations wherein CEOs or chairpersons concurrently hold multiple roles within distinct tiers of government agencies, the value assigned to this variable corresponds to the utmost highest hierarchical level among all occupied positions (Wang, Jia, & Zhang, Reference Wang, Jia and Zhang2021). In addition, a firm’s poverty alleviation strategy can be influenced if the founder serves as the CEO, as they typically hold greater decision-making authority. To account for this, the variable Founder CEO was controlled. This variable was assigned a value of 1 if the CEO is the founder and 0 otherwise.
In addition, we included four provincial-level variables to control for local government influences. Specifically, we controlled for the logarithmic regional GDP per capita (Provincial per capita GDP, ln-transformed), the local government’s budgetary preferences for education expenditure (Provincial education preference), social security expenditure (Provincial social security preference), and environmental protection expenditure (Provincial social environment preference). The latter three variables were measured as the respective budgetary expenditures divided by the total general budget expenditure of the local government for the year.
Finally, we considered year-, industry-, firm-, and provincial-fixed effects to control for common temporal trends in poverty alleviation and disparities across firms, industries, and provinces over the years (Xu & Ma, Reference Xu and Ma2022). To mitigate potential reverse causality concerns, the independent variable and all control variables were lagged by 1 year. We scaled certain control variables by a factor of 10 to improve the clarity of the coefficient interpretations. These variables included SOE, ROA, firm age, revenue growth, firm size, CEO education, CEO gender, CEO duality, and political connection level.
Research Model
The following regression equations were utilized for our analysis:
\begin{align}{\text{Generic poverty alleviation}} &= {{{\beta }}_{\text{0}}} +{{{\beta }}_{\text{1}}} \times {\text{CEO early}}\text{-}\text{life poverty trauma exposure} \nonumber\\
&\quad + {{{\beta }}_{\text{2}}} \times {\text{Controls}} + \varepsilon \end{align}
\begin{align}{\text{Strategic poverty alleviation}} &= {{{\beta }}_{\text{0}}} + {{{\beta }}_{\text{1}}} \times {\text{CEO early}}\text{-}\text{life poverty trauma exposure} \nonumber\\
&\quad + {{{\beta }}_{\text{2}}} \times {\text{Controls}} + \varepsilon \end{align} where Generic poverty alleviation and Strategic poverty alleviation are our two dependent variables, CEO early-life poverty trauma exposure is the independent variable, Controls represents all our control variables, and 
${{\varepsilon }}$ is the error term.
Following previous studies (e.g., Cao, Jiang, & Wang, Reference Cao, Jiang and Wang2024; Dineen, Duffy, Henle, & Lee, Reference Dineen, Duffy, Henle and Lee2017), we first ran each of our two models through fixed-effect regression, clustering the standard errors at the firm level to control for within-firm correlations to test our H1 and H2, respectively. Given the consistency in the set of exploratory variables across the two equations and the correlation between the two dependent variables, suggestive of a common unobserved disturbance, we employed seemingly unrelated estimations (the suest command in STATA 15.0) to jointly estimate the two equations (Li, Hernandez, & Gwon, Reference Li, Hernandez and Gwon2019; Rhee & Fiss, Reference Rhee and Fiss2014; Sharapov & Ross, Reference Sharapov and Ross2023). The suest postestimation test is an advanced statistical method that facilitates the comparison of coefficients across multiple models by accounting for correlated errors. This procedure consolidates the estimation results into a single, simultaneous covariance matrix, enabling robust hypothesis testing both within and between models. It is particularly suitable for testing our H3. Following previous studies (Ertug, Yogev, Lee, & Hedstrom, Reference Ertug, Yogev, Lee and Hedstrom2016; Li et al., Reference Li, Hernandez and Gwon2019), after conducting the suest procedure, we used a test command in STATA (Rhee & Fiss, Reference Rhee and Fiss2014) to compare the effects of CEOs’ early-life poverty trauma exposure on the firm’s generic poverty alleviation and strategic poverty alleviation.
Results
Table 1 presents the descriptive statistics and the correlation matrix of key variables in the main examination. It shows that there were high correlations between some explanatory variables, such as the positive correlation between CEO duality and Founder CEO (0.580) and the positive correlation between Founder CEO and Manager share ratio (0.556). We thus investigated whether a potential multicollinearity problem exists in our regression model by computing variance inflation factors (VIFs). The maximum VIF in any of our regression models was 2.18 (attributed to Founder CEO), and the mean VIF was 1.37, substantially below the rule-of-thumb cutoff of 5 for regression models. Therefore, multicollinearity was not a critical issue in our regression models.
Descriptive statistics and correlation matrix
Notes: Number of firm-year observations = 1246. Correlations greater than 0.056 or less than −0.056 are significant at 5% level.
a We present the raw values (in millions) in this table. Due to the difference in scale between generic and strategic poverty alleviation, we standardized these variables in our regression analyses and then compared the effects of CEO early-life poverty trauma exposure on firm generic poverty alleviation and that on strategic poverty alleviation.
The main results of the analysis are summarized in Table 2. Models 1 and 2 included all the control variables to predict the generic and strategic poverty alleviation, respectively. Model 3 underscored the main effect of CEOs’ early-life poverty trauma exposure on the firm’s generic poverty alleviation. The analysis revealed that the impact of CEOs’ early-life poverty trauma exposure on the firm’s generic poverty alleviation was significantly positive (β = 0.142, p < 0.01), thus supporting our H1. Model 4 indicated a positive and significant relationship between CEOs’ early-life poverty trauma exposure and the firm’s strategic poverty alleviation (β = 4.583, p < 0.01), supporting our H2. We further compared coefficients on CEOs’ early-life poverty trauma exposure in Model 3 and that in Model 4. The comparison showed that the effect of CEOs’ early-life poverty trauma exposure on strategic poverty alleviation was significantly greater than that on generic poverty alleviation (β 1 = 0.142, β 2= 4.583, p = 0.005), thereby supporting our H3.
Regressions predicting generic poverty alleviation and strategic poverty alleviation
+ Note: t-statistics in parentheses, p < 0.10, *p < 0.05, **p < 0.01.
Supplementary Tests
Post hoc analysis of the mediating effect of emphasis on resource efficiency
In our previous discussions, we argued that a mindset focused on resource efficiency drives these CEOs to prefer strategic initiatives, which are typically more effective and impactful, over generic ones. In this section, we conducted a test to examine the mediating effect of emphasis on resource efficiency on the relationship between CEOs’ early-life poverty trauma exposure and firms’ relative emphasis on strategic over generic involvement in poverty alleviation. CEOs with early-life poverty trauma possess a deep understanding of poverty’s entrenched and complex nature. Pertinently, previous research has suggested that those who have experienced poverty tend to sharpen their focus on resource efficiency, especially in resource-constrained environments (Evans, Reference Evans2004; Sheehy-Skeffington & Rea, Reference Sheehy-Skeffington and Rea2017). Therefore, such exposure enables CEOs to identify and prioritize truly essential resources for poverty alleviation, resulting in more effective resource allocation decisions (Gupta et al., Reference Gupta, Briscoe and Hambrick2018), i.e., emphasis on resource efficiency. Furthermore, this mindset, aligned with the economic principle of scarcity, emphasizes optimal resource utilization, serving as a critical element in strategic decision-making (Krautkraemer, Reference Krautkraemer1998). Considering the above arguments, we propose that early-life poverty exposure enhances CEOs’ emphasis on resource efficiency, which in turn strengthens their engagement in strategic poverty alleviation efforts over generic involvement. In other words, emphasis on resource efficiency mediates this relationship.
To compute emphasis on resource efficiency, we employed the Word2Vec program to develop a lexicon of resource efficiency-related terms, which had been widely used by previous studies (Guo et al., Reference Guo, Sengul and Yu2021; Harrison et al., Reference Harrison, Thurgood, Boivie and Pfarrer2019; Miric, Jia, & Huang, Reference Miric, Jia and Huang2023). This method accounts for the context in which words appear, thereby offering a comprehensive and objective measure of semantic similarity among the terms.
We measured Emphasis on resource efficiency using the machine learning technique. First, five representative seed words – ‘Efficiency’, ‘Productivity’, ‘Cost reduction and efficiency improvement’, ‘Efficiency enhancement’, and ‘Work efficiency’ – were compiled according to suggestions of academic and industry experts. We consulted one accounting professor and one MBA student to generate an initial list of words related to resource efficiency emphasis. Then, we engaged two MBA students to identify words from this list that best reflect firms’ focus on resource efficiency emphasis. If there was any disagreement among them, they would discuss and reach a consensus, and we finally got these five seed words. The second stage involved using the Word2Vec machine learning model to expand this seed word set. This technique was used to represent natural language words as mathematical vectors, which were then mapped in higher-dimensional space to determine word similarity (Harrison et al., Reference Harrison, Thurgood, Boivie and Pfarrer2019; Mikolov, Sutskever, Chen, Corrado, & Dean, Reference Mikolov, Sutskever, Chen, Corrado and Dean2013). After eliminating repeated words, the set was refined by sampling words and comparing them against firm efficiency descriptions in annual reports for consistency. This process culminated in a comprehensive set of 20 words denoting emphasis on resource efficiency (including ‘Management Efficiency’, ‘Labor Productivity’, ‘Utilization Efficiency’, ‘Enhance Efficiency’, ‘Cost Reduction and Efficiency Improvement’, ‘Administrative Efficiency’, ‘Manual Efficiency’, ‘Work Efficiency’, ‘Labor Efficiency’, ‘High Yield’, ‘Productivity’, ‘Optimization for High Efficiency’, ‘High Efficiency’, ‘Highly Efficient’, ‘Efficiency’, ‘Value Creation’, ‘Labor Effectiveness’, ‘Efficient’, ‘Efficiency Enhancement’, and ‘Efficiency Improvement’). Subsequently, we calculated the frequency of these terms appearing in the Management Discussion and Analysis section of annual corporate reports, as the information disclosed in this part had been found to reflect managerial intents (e.g., Feldman, Govindaraj, Livnat, & Segal, Reference Feldman, Govindaraj, Livnat and Segal2010). We then log-normalized these frequencies to measure emphasis on resource efficiency. The mean value and standard error for this variable were 1.241 and 0.785, respectively.
We further constructed a ratio measure as our dependent variable to capture the relative resource allocation between strategic and generic poverty alleviation. This variable, named as Allocation of poverty alleviation, was operationalized as the value of strategic poverty alleviation divided by the sum of generic poverty alleviation plus one (plus one, ln-transformed). The mean value and standard error for this variable were 0.605 and 1.722, respectively. We then controlled for the logarithmic total poverty alleviation engagement (Total poverty alleviation engagement) in our model.
Table 3 shows the results of the three-step regression method. Model 1 in Table 3 showed that there was a significant positive direct relationship between CEO early-life poverty trauma exposure and Allocation of poverty alleviation (β = 48.379, p < 0.01), indicating that CEOs’ early-life poverty trauma exposure has a greater positive effect on firm strategic poverty alleviation than general poverty alleviation. Model 2 in Table 3 showed the significant effect of CEO early-life poverty trauma exposure on Emphasis on resource efficiency (β = 7.043, p < 0.05). In Model 3 of Table 3, the relationship between Emphasis on resource efficiency and Allocation of poverty alleviation (β = 0.129, p < 0.05) was significantly positive, and the relationship between CEO early-life poverty trauma exposure and Allocation of poverty alleviation was still significantly positive (β = 47.468, p < 0.01). Consistent with our expectation, the coefficient on CEO early-life poverty trauma exposure in Model 3 was smaller than that in Model 1. We finally computed the bootstrapped confidence intervals (500 iterations) and the results indicated a significant mediation effect of firm efficiency orientation (95% CI [1.390, 1.467]).
Supplementary analysis: the mediating role of emphasis on resource efficiency
+ Notes: t-statistics in parentheses, p < 0.10, *p < 0.05, **p < 0.01.
Alternative measurements of the dependent variable
To mitigate concerns regarding the broad classification of generic and strategic involvements in poverty alleviation, we examined and compared the effects of CEOs’ early-life poverty trauma exposure on healthcare poverty alleviation (a representative type of generic poverty alleviation) and industrial poverty alleviation (a representative type of strategic poverty alleviation). It is noteworthy that industrial poverty alleviation, being a crucial poverty alleviation strategy in China, had affected approximately 30 million impoverished individuals (Saci, Reference Saci2023). Models 1 and 2 in Table 4 displayed the consistent results of this robustness test.
Regression results for robustness tests 1–4
+ Notes: t-statistics in parentheses, p < 0.10, *p < 0.05, **p < 0.01.
We further classified environmental protection poverty alleviation efforts in heavily polluting industries as strategic poverty alleviation. Models 3 and 4 in Table 4 showed the consistent results of this robust test. Additionally, when we excluded educational poverty alleviation from strategic poverty alleviation, the results in Models 5 and 6 of Table 4 still supported our hypotheses.
Moreover, we conducted a further analysis by reclassifying generic and strategic poverty alleviation programs based on their relevance to the firm’s business operations. Specifically, in the first step, we extracted textual descriptions of the firms’ business operation scopes from their annual reports. We then manually coded each firm’s poverty alleviation practices as either generic or strategic involvement, based on their alignment with the business scope description. In the second step, we calculated our new dependent variables using the same method outlined in the above method part, but based on these newly coded categories. This measurement can capture how the poverty alleviation practice is integrated with the firm’s operation or how it has been closely monitored by the firms, who would then use their operational strength to enable the implementation. The results in Models 7 and 8 of Table 4 were consistent with the main test.
Effect on total engagement in poverty alleviation
Our arguments suggested that CEOs’ early-life poverty trauma exposure drove the emphasis on resource efficiency and subsequently resulted in a firm’s relative emphasis on strategic involvement in poverty alleviation over generic involvement. We conducted a further analysis to examine whether such exposure would affect the overall poverty alleviation involvement. Our findings shown in Model 1 of Table 5 suggested that CEOs with early-life poverty trauma exposure would engage in more poverty alleviation activities. Furthermore, taking the total amount of poverty alleviation into consideration, we measured generic poverty alleviation as the amount of generic poverty alleviation divided by the total amount of poverty alleviation alternatively and measured strategic poverty alleviation as the amount of strategic poverty alleviation divided by the total amount of poverty alleviation alternatively. Results presented in Models 2 and 3 of Table 5 indicated that CEOs with early-life poverty trauma would allocate more resources to strategic poverty alleviation activities than generic poverty alleviation programs.
Regression results for robustness tests 5–9
+ Notes: t-statistics in parentheses, p < 0.10, *p < 0.05, **p < 0.01, ***p < 0.001.
Effects on poverty alleviation versus poverty alleviation unrelated CSR (PA-unrelated CSR)
In this section, we compared the effects of CEOs’ early-life poverty trauma exposure on poverty alleviation and PA-unrelated CSR. Specifically, we obtained firms’ CSR data from the Chinese Research Data Service Platform database, which had been widely used by previous studies (e.g., Yang, Orzes, Jia, & Chen, Reference Yang, Orzes, Jia and Chen2021; Zhang, Chen, & Jia, Reference Zhang, Chen and Jia2023). It encompasses multiple measures of strengths and concerns across six ESG dimensions: charity, corporate governance, diversity, employee relations, environment, and product. Each dimension is evaluated using various items, with each item scored as either ‘1’ (indicating the firm has exhibited this strength or concern) or ‘0’ (indicating otherwise). We calculated the CSR score for each dimension by summing the scores of the strengths items and subtracting the scores of the concerns items within that dimension. We excluded the charity dimension due to its high relevance to poverty alleviation efforts, such as engagement in education support, employment promotion, and local economic development. We also excluded the corporate governance dimension due to its low association with social activities. Additionally, we excluded the sub-dimension of diversity named ‘innovative human resource project’, as it aims to help disabled people and overlaps with generic poverty alleviation.
Then, to compare the impacts of CEOs’ early-life poverty trauma exposure on PA-unrelated CSR and poverty alleviation efforts, we standardized both the firm’s PA-unrelated CSR score and the total amount of engagement in poverty alleviation. The results displayed in Models 4 and 5 of Table 5 revealed that the effect on poverty alleviation is greater than that on PA-unrelated CSR.
Controlling for potential peer effects
To test the potential influence of peer effect, we conducted two additional robustness tests. First, we included the industry average levels of generic and strategic poverty alleviation in our regression model. The consistent results were presented in Models 6 and 7 of Table 5. Second, we adjusted the values of generic and strategic poverty alleviation by subtracting the industry average for each and then standardized these two variables. The consistent results were shown in Models 8 and 9 of Table 5.
Alternative measurements of the independent variable
We explored alternative measurements for our independent variable. In the main analysis, we defined our treatment group as CEOs who were 7–11 years old during 1959–1961, and our control group as those born during 1962–1964. To address concerns that our independent variable may be sensitive to the age period choice, we redefined the treatment groups as CEOs who were 7–12, 7–13, and 7–14 years old during 1959–1961, respectively. The results were shown in Models 1–6 of Table 6 and were consistent with our expectations. We further constructed our control group to include CEOs who were born within four years after the famine (1962–1965). Results shown in Models 7 and 8 of Table 6 were robust. It was noted that the independent variables for these four robustness checks are labeled as CEO early-life poverty trauma exposure 1, 2, 3, and 4, respectively. Additionally, we used the dummy variable CEO poverty early-life experience as our independent variable to differentiate between CEOs with and without poverty experience regarding their engagement in poverty alleviation. The qualitatively consistent results of this analysis were presented in Models 9 and 10 of Table 6.
Regression results for robustness tests 10–14
+ Notes: t-statistics in parentheses, p < 0.10, *p < 0.05, **p < 0.01.
Considering the effect of chairman poverty experience
We considered the possibility that the chairman may serve as the ‘first-in-command’ and wield more power than the CEO in some firms. To account for this situation, we used the criterion of managerial ownership to judge the leader role in a firm because the ownership is positively associated with the hierarchical influence of executives (Denis, Denis, & Sarin, Reference Denis, Denis and Sarin1997; Morck, Shleifer, & Vishny, Reference Morck, Shleifer and Vishny1988). It was thus reasonable to expect that when the chairman’s ownership was higher than that of the CEO, the chairman might have greater power and influence on the firm’s poverty alleviation participation. We thus conducted a robustness check by substituting the CEO’s early-life poverty trauma exposure, education, and gender with those of the chairman when the chairman’s ownership stake was higher than that of the CEO. The results were presented in Models 1 and 2 of Table 7.
Regression results for robustness tests 15–18
+ Notes: t-statistics in parentheses, p < 0.10, *p < 0.05, **p < 0.01, ***p < 0.001.
a Given the reduced sample size and the risk of overfitting due to the inclusion of multiple variables in our models (Heimeriks, Schijven, & Gates, Reference Heimeriks, Schijven and Gates2012; Tong, Wang, & Xia, Reference Tong, Wang and Xia2020), we streamlined our regression model. Therefore, we included only the independent variable, Famine experience change, along with three basic control variables: ROA change, revenue growth change, and firm size change.
b As this model is a probit regression model, therefore, firm fixed effect was not controlled.
Excluding firms designated as special treatment
We excluded firms that had been designated as ‘Special Treatment’ (ST or *ST) by stock exchanges, as these entities were subject to a range of trading and financial restrictions (Wang, Wijen, & Heugens, Reference Wang, Wijen and Heugens2018). In China, any firm that suffered losses for two consecutive fiscal periods would be accorded this special treatment status. This status would subsequently limit its daily price fluctuations to 5% and heighten its risk of being delisted if it reported losses for 4 consecutive years, accompanied by negative net assets for 3 consecutive years (Quan, Ke, Qian, & Zhang, Reference Quan, Ke, Qian and Zhang2023). Models 3 and 4 in Table 7 presented the results of this robustness check. All the results from these robustness tests remained largely congruous with those from the main test.
Heckman two-stage regression model
Given that our sample included only firms that have disclosed their CEOs’ birthplace, selection bias might exist (Certo, Busenbark, Woo, & Semadeni, Reference Certo, Busenbark, Woo and Semadeni2016). To address this concern, we applied Heckman’s (Reference Heckman1979) two-stage approach. In the first stage, we estimated a probit model predicting the probability of sample inclusion, with the dependent variable (labeled as CEO birthplace disclosure) coded as 1 for firms in our sample and 0 for those not included. The industry average level of CEO birthplace disclosure was used as an instrumental variable, which influenced whether a firm disclosed the CEO’s birthplace but had no direct influence on the firm’s poverty alleviation strategy. The significant positive effect of this industry average disclosure on CEO birthplace disclosure (β = 4.027, p < 0.001, Model 5 of Table 7), combined with its lack of impact on either Generic poverty alleviation or Strategic poverty alleviation, further supported the reliability of this instrumental variable. In the second stage, we included the inverse Mills ratio derived from the first-stage probit model to control for potential sample selection bias. The results from the second-stage regression models (Models 6 and 7 of Table 7) were consistent with those from the main analysis.
CEO turnover analysis
We finally conducted a robustness test to account for the effect of CEO turnover. Specifically, we examined whether changes in CEOs’ early-life poverty trauma exposure influenced a firm’s generic and strategic poverty alleviation efforts. We restricted our sample to CEO turnover events within our sample period, reducing the number of observations to 145. We captured the effect of Δ CEO famine experience on Δ firm poverty alleviation by constructing a change variable, Famine experience change. This variable was assigned a value of −1 if the predecessor CEO experienced famine during childhood while the successor CEO did not. Conversely, it was assigned a value of 1 if the predecessor CEO did not experience famine during childhood but the successor CEO did. The variable was assigned a value of 0 if both the predecessor and successor CEOs either experienced or did not experience famine during childhood. Due to the small sample size and potential overfitting issues (Heimeriks, Schijven, & Gates, Reference Heimeriks, Schijven and Gates2012; Tong, Wang, & Xia, Reference Tong, Wang and Xia2020), we only included the independent variable Famine experience change and three basic control variables (ROA change, Revenue growth change, and Firm size change) in the model. We then regressed these variables against standardized changes in generic poverty alleviation and strategic poverty alleviation to compare the effects of CEO famine experience change on these outcomes. The results shown in Models 8 and 9 of Table 7 were still robust.
Discussion
The literature on imprinting theory suggests that individuals’ behavioral preferences and mindsets are subject to environmental characteristics, which ‘reflect prominent features of the environment, and these characteristics persist despite significant environmental changes in subsequent periods’ (Marquis & Tilcsik, Reference Marquis and Tilcsik2013: 199). Childhood stands out as a critical sensitive period in shaping one’s traits (Han, Chi, & Zhou, Reference Han, Chi and Zhou2022; Kish-Gephart & Campbell, Reference Kish-Gephart and Campbell2015). Previous studies have identified that CEOs’ early-life poverty trauma exposure plays a non-negligible role in shaping CEOs’ empathy tendencies and risk preferences, which in turn markedly influence firms’ prosocial strategies and risk-taking behaviors (e.g., Chang et al., Reference Chang, He and Wang2021; Choi et al., Reference Choi, Shin and Kim2023; Han et al., Reference Han, Chi and Zhou2022; Zhang, Reference Zhang2017). Based on this, our study aims to explore how CEOs’ early-life poverty trauma exposure impacts firms’ involvement in distinct poverty alleviation initiatives. We classify poverty alleviation programs into generic and strategic ones, and we find that CEOs with larger early-life poverty trauma exposure are more likely to engage in these two types of poverty alleviation. In addition, we further observe that CEOs’ early-life poverty trauma exposure has a greater positive impact on a firm’s strategic involvement in poverty alleviation than on its generic involvement because of the emphasis on resource efficiency.
Our study makes several contributions to the extant literature. First, our study contributes to the literature on the relationship between CEO early-life experience and social practice engagement. Previous research primarily focused on how early-life experience matters for commitment to the well-being of others and the formation of a sense of responsibility (e.g., Chang et al., Reference Chang, He and Wang2021; Choi et al., Reference Choi, Shin and Kim2023; O’Sullivan et al., Reference O’Sullivan, Zolotoy and Fan2021). For instance, CEOs who experienced famine, poverty, or natural disasters during their childhood tend to make significant philanthropic contributions and show enhanced CSR performance (Choi et al., Reference Choi, Shin and Kim2023; Han et al., Reference Han, Chi and Zhou2022; Kahhali, Reference Kahhali2022; Liu et al., Reference Liu, Zhang and Zhang2024). Our study offers new insights into the literature on managerial decision-making by emphasizing how CEOs’ differentiated preferences and inclinations shape their strategic choices. Specifically, in the context of poverty alleviation, we align with previous research suggesting that CEOs with early-life poverty exposure are more likely to engage in both types of poverty alleviation efforts, reflecting a general tendency toward prosocial practices. Furthermore, we propose that early-life poverty exposure may foster a strategic mindset centered on resource efficiency. This mindset prompts CEOs to prioritize the optimal use of resources to generate meaningful social impact, indicating a shift from traditional charitable giving to a more purpose-driven and highly efficient approach to poverty alleviation. This emphasis on resource efficiency reflects a pragmatic and strategically oriented commitment to social investment, wherein resource allocation is carefully managed to ensure substantial and enduring outcomes. By highlighting the role of early-life experiences in shaping strategic approaches to prosocial activities, our study adds depth to strategic management literature, underscoring how early-life experience can influence CEOs’ strategic priorities and decision-making framework in socially responsible initiatives.
Second, our study advances the literature on the substantiveness of social practices. Previous studies mainly focused on the moral lens of the difference between symbolic and substantive CSR, as well as their respective outcomes. If there is a gap between social practice communication and implementation, it is regarded as a symbolic prosocial practice and may trigger criticism of being hypocritical. For instance, symbolic engagement in prosocial practice may generate reputational damage (Graafland & Smid, Reference Graafland and Smid2019), erode trust among key stakeholders (Barnett, Reference Barnett2007), and impair customer satisfaction (Ioannou, Kassinis, & Papagiannakis, Reference Ioannou, Kassinis and Papagiannakis2023). In our context of poverty alleviation, we further propose another lens to understand the extent of substantiveness of social practice. We classify poverty alleviation into generic involvement, characterized by short-term financial support and limited resource efficiency (e.g., donations and relief aid), and strategic involvement, which more closely aligns the firm’s resources with the needs of disadvantaged communities, fostering sustainable, long-term impacts that enable community self-sufficiency. Thus, while previous research has primarily focused on the differences between symbolic and substantive social practice, our study introduces a vital dimension by evaluating the extent of substantiveness through the lens of resource efficiency and instrumental value for both society and the firm itself. We further propose that CEOs exposed to early-life poverty trauma may prioritize strategic poverty alleviation over generic ones. Even though both forms of engagement are substantive, CEOs with early-life poverty experience are better able to understand the root causes of poverty and the needs of impoverished populations, enabling them to adopt more resource-efficient strategic poverty alleviations, which seem more substantive than generic ones.
Third, our study aligns with the themes of grand challenges and the UN SDGs by focusing on poverty alleviation in the context of transforming economies, with particular attention to China. Unlike previous studies that typically address general CSR activities, our research examines a context-specific aspect of CSR in China: poverty alleviation involvement. This focus is especially pertinent given China’s prioritization of poverty alleviation as a significant national objective. China’s unique context, characterized by rapid economic growth alongside persistent regional inequalities, provides a valuable setting for exploring poverty alleviation within the transforming economies. By investigating poverty alleviation efforts in China, we uncover the complex, multifaceted nature of poverty alleviation initiatives, which range from short-term relief efforts to long-term capacity-building programs targeting structural poverty. Furthermore, we highlight the critical impact of CEOs’ emphasis on a resource efficiency mindset in enhancing the substantive effectiveness of poverty alleviation policies. Our findings illustrate that effective poverty alleviation requires not only the demands of the institutional environment but also a high degree of alignment with individual psychological factors, thereby maximizing the value of poverty alleviation efforts. Our research thus offers a nuanced understanding of the challenges and strategies associated with achieving the ‘No Poverty’ goal, which also provides implications for other transforming economies, which face similar structural challenges.
Our study has several practical implications for both firms and policymakers. First, this study differentiates between generic and strategic poverty alleviation. Thus, from the external stakeholder management perspective, firms might consider aligning their poverty alleviation or broader social practice with core business scopes. For instance, firms could invest in skill development or vocational training programs within communities, which not only contribute to sustainable poverty reduction but also provide a skilled labor force for the firm, thereby enhancing both economic and social outcomes. From the internal corporate management perspective, while traditional executive selection processes typically highlight an individual’s professional accomplishments, it is crucial for firms to consider a candidate’s mindset resulting from early-life experiences. Although these experiences are rarely detailed in a CEO’s resume, the executive selection process provides an opportunity to discover these valuable life stories. Our findings suggest that individuals with early-life poverty trauma exposure are inclined toward altruism and demonstrate a strong emphasis on resource efficiency. We should acknowledge that these attributes may affect a firm’s approach to managing social issues and even other corporate policies.
As for policymakers, there is potential for collaboration between the government and firms in the area of poverty alleviation. Policymakers can encourage more firms to participate in poverty alleviation projects – especially those with strategic significance and substantial impact – by providing incentive policies, tax benefits, and other supportive measures. For firms that hold a leading position in key areas requiring targeted efforts, additional financial support or policy incentives may be offered to maximize their poverty alleviation efforts in these critical areas. Furthermore, our study highlights the importance for policymakers to understand that the interaction between government and firms should be not only at the resource level but also at the cognition level. In particular, regarding poverty alleviation initiatives, firm leaders may take different approaches to addressing poverty alleviation based on their exposure to poverty trauma experiences. It is crucial for governments to align their policies with the preferences and motivations of firm leaders to foster effective partnerships and enhance the efficiency of poverty alleviation efforts. This approach not only advances the goals of poverty alleviation but also fosters a cooperative environment that maximizes the strengths and resources of both the government and firms.
Limitations and Future Research
This study inevitably has several limitations. First, in line with prior research in the upper echelon literature, we rely on demographic information such as birth year and birth province to gauge CEOs’ early-life poverty trauma exposure. This approach, while useful, has its drawbacks. Specifically, our sample may be subject to bias due to missing data on CEO birthplace information, which can affect the comprehensiveness of our analyses. Additionally, demographic proxies might not fully capture the nuanced and personal nature of early-life poverty trauma exposure. Second, although we have utilized machine learning techniques to capture the emphasis on resource efficiency and have conducted mediation analysis, this method is not immune to flaws. The ideal solution is to understand individuals’ mindsets and psychological traits by utilizing the experience method or interviews to provide richer insights into the processes driving their strategic decisions in poverty alleviation initiatives. Third, the research focus of our study is to investigate the relative emphasis on involvement of strategic poverty alleviation over that of generic poverty alleviation. We believe that another promising venue for scholars is to further examine the differentiation approach to poverty alleviation. It can be a future direction for future research to complement the literature on strategy differentiation or uniqueness. Fourth, our investigation is based on the context of Chinese firms; the chairman may play a more important role than the CEO in some firms. We have conducted a test by relying on the ownership hierarchy between the chairman and CEO. However, we cannot fully identify whether the CEO or the chairman plays a more critical role in decision-making, as they may be responsible for different types of strategies and activities related to business operations. We acknowledge this limitation, which opens an avenue for future research. Moreover, the focus on the Chinese context may limit the generalizability of our findings. Future research should be cautious about this limitation when generalizing our findings to other institutional settings. Relatedly, it’s very valuable to conduct comparative studies in other economies to assess how CEOs’ early-life poverty trauma exposure influences the engagement of firms in different types of prosocial activities across countries or institutions.
In conclusion, our study develops a framework to elucidate how CEOs with early-life poverty trauma exposure prioritize between generic and strategic involvement in poverty alleviation. This framework paves the way for future research to deepen our understanding of the intricate relationship between CEOs’ early-life experiences and prosocial decisions.
Data availability statement
The data that support the findings of this study are available from the corresponding author upon reasonable request.
Acknowledgements
This study was funded by grants from the National Natural Science Foundation of China (Nos. 72102003, 72102184, 72091310, and 72091314) and the Humanities and Social Science Project, Ministry of Education of China (No. 21YJC630128).
Yongzhi Du (2018100728@mail.nwpu.edu.cn) is an Assistant Professor of management at the School of Business, Harbin Institute of Technology, Harbin, Heilongjiang, China. His research interests include corporate social responsibility and corporate Internationalization strategy. He has published in Corporate Social Responsibility and Environment Management and Chinese Management Studies.
Li Tong (litong2020@gsm.pku.edu.cn) is an assistant professor of management at the Guanghua School of Management, Peking University, China. His research interests include corporate social responsibility and corporate M&A strategy. He has published in the Academy of Management Journal, Strategic Management Journal, and Journal of Business Ethics.
Wenxin Wu (wuwx0606@gmail.com) is an assistant professor of management at the Advanced Institute of Business, Tongji University, China. Her research interests include corporate social responsibility and corporate governance. She has published in the Strategic Management Journal.
Chenjing Wang (2101111023@gsm.pku.edu.cn and chenjingwxy@stu.pku.edu.cn) is a PhD candidate in strategic management at Guanghua School of Management, Peking University, China. Her research interests include international business, corporate social responsibility, and corporate governance.
