Internal CSR and blue-collar workers’ attitudes and behaviors in China: a combination of a cross-sectional study and a field experiment

While a positive relationship between corporate social responsibility (CSR) and employee outcomes has been widely researched, most studies have been conducted in a Western context and for white-collar employees. Thus, little is known about Chinese blue-collar workers’ reactions to CSR. In two studies, we investigated the outcomes of internal CSR for blue-collar workers in the Chinese manufacturing industry. In the cross-sectional study, we found that perceived internal CSR is positively related to job satisfaction, trust in management, and perceived supervisory support. To further disentangle cause-and-effect relationships, we conducted a field experiment. The results show that internal CSR information increases performance quantity and lowers complaints, but also lowers performance quality and does not affect workers’ voice behavior. Our results further reveal that workers’ attitudes toward ethics and social responsibility are an important moderator which strengthens some of the hypothesized effects.


Introduction
Despite growing scholarly attention to the economic outcomes of corporate social responsibility (CSR), little is known about the effects of CSR in emerging countries in general and in China in particular (Barakat et al., 2015;Shang et al., 2022;Zhao et al., 2019). This situation is somewhat surprising given that different stakeholders increasingly put firms in China under pressure to pay more attention to socially responsible behavior (Dang et al., 2020;Li et al., 2016;Ng et al., 2022;Parsa et al., 2021;Yin, 2017). The relevance of internal CSR, i.e. activities that improve the well-being of employees (De Roeck & Maon, 2018;Jamali et al., 2020), has been amplified over the past years as Western companies nowadays require their Chinese suppliers to adhere to social and labor standards (Ding et al., 2021;Yin & Zhang, 2012). In addition, the Chinese government has been exerting considerable pressure on companies to practice (internal) CSR through various regulatory initiatives and the introduction of the social credit system for businesses (De Jonge, 2021;Ervits, 2021;Hofman et al., 2017).
In light of the growing pressure for Chinese companies to assume responsibility for the well-being on their employees, it is fruitful to investigate the work-related outcomes of internal CSR in this setting (Xiao et al., 2020), especially against the background that the Chinese manufacturing industry is regularly being accused of the prevalence of poor working conditions (Chen et al., 2017;Sznajder et al., 2022), which threatens the long-term success of Chinese factories (Lin-Hi et al., 2019). Therefore, it is valuable to address the question of whether internal CSR can promote firms' and workers' interests at the same time.
While employee-level outcomes of internal CSR have received growing scholarly attention over the past years, existing empirical studies are centered on white-collar and educated employees from Western, industrialized countries (Gond et al., 2017;Wang et al., 2020b). However, since the meaning of CSR and its individual-level outcomes are strongly affected by national culture (Chen et al., 2020;Hong et al., 2021;Mueller et al., 2012), it is necessary to replicate existing findings in different contexts and with alterations in methods and subjects. This circumvents the pitfalls associated with the adoption of a "universalist" approach by which Western-based theoretical and practical insights (about CSR) are simply transferred to developing countries (Banerjee et al., 2021;Cooke, 2018;Ma & Bu, 2021).
The present paper aims to contribute to a better understanding of internal CSR's effects on work-related attitudes and behaviors in the specific context of the Chinese blue-collar manufacturing sector. To this end, we draw on social exchange theory (Blau, 1964) and person-organization (PO) fit theory (Kristof, 1996). We conducted two studies in a garment factory in China. In a survey study, we investigated the link between workers' perceptions of internal CSR and three work-related attitudes that capture workers' relationships with different organizational facets and actors (i.e., the job itself, top management, and direct supervisors): job satisfaction, trust in management, and perceived supervisory support. In the second study, we conducted a field experiment to evaluate the causal effects of internal CSR on the following behaviors: task performance, voice behavior, and complaints. In both studies, we analyzed whether workers' perceived importance of the role of ethics and social responsibility (PRESOR importance) moderates the hypothesized relationships. Figure 1 summarizes our research framework.
The present article responds to the recent calls to conduct CSR research outside the Western world (Jamali et al., 2020;Khan et al., 2021;Tilt, 2016) and, by adding new, context-specific insights, contributes to growing empirical research on the impact of (internal) CSR on employee attitudes and behaviors (for a review, see Gond et al., 2017;Wang et al., 2020aWang et al., , 2020b. Previous studies on the relationship between CSR and similar employee-level outcomes have been mostly cross-sectional field studies and are therefore limited in regards to the derivation of cause-effect relationships (Aguinis & Glavas, 2012;Jones et al., 2019). By carrying out a field experiment, we rule out possible common method biases and are able to demonstrate causal relationships. As an extension of previous research, we also look at outcomes that have not been considered so far and address the moderating effect of workers' PRESOR importance which has been previously shown to be a relevant personal attitude in CSR studies (Tian & Suo, 2021;Zhao et al., 2021). Our findings provide several theoretical and practical implications, including recommendations for CSR strategy and communication tailored to the Chinese context.

Internal CSR in the Chinese blue-collar context
While the concept of CSR is relatively new to China compared to the Western world (Kolk & Tsang, 2017;Ma & Bu, 2021), its importance has been steadily growing over the last years (Lu, 2018;Xu & Liu, 2020). Nowadays, an increasing number of Chinese private and state-owned companies are actively involved in CSR and several institutions, such as Hexun and Runlin Global's Rankings, rate the CSR performance of Chinese companies (Zhong et al., 2019). Chinese firms also increasingly report about their CSR activities and participate in global CSR initiatives. For example, between 2008 and 2019, the number of issued CSR reports in China has increased by more than fivefold (Khalid et al., 2022a) and the membership of Chinese companies in the United Nations Global Compact grew from 10 in 2005 to more than 500 in 2021 (Kim & Koo, 2022).
Given China's unique historical, socio-economic, political, and cultural background, CSR exhibits a number of unique characteristics in this country (Ervits et al., 2021;Ma & Bu, 2021;Zhang et al., 2019). A distinguishing factor compared with the Western world lies in the country's "top-down" (Parsa et al., 2021), "state-centric" (Kolk & Tsang, 2017) approach to CSR, with the Chinese central government assuming a proactive role in the promotion of responsible business practices through legislation (Banik & Lin, 2019;Ng et al., 2022;Tang et al., 2018;Wang et al., 2018b;Zhang et al., 2021b). Especially since 2006, following the introduction of the concept of "Harmonious Society", CSR has become a national and strategic priority for the Chinese government on the way to solving various social issues that have arisen due to the country's focus on economic growth, such environmental pollution, inequality in regional development (Bu et al., 2022;Zhang et al., 2022b). Since then, the government has introduced a number of laws and regulations, which belong to the dominant driving forces for the development and implementation of CSR in China (Li & Appiah-Konadu, 2021). For example, in 2006, China passed the revised Company Law specifying business' obligations to behave in a socially responsible fashion which market the first time that the government recognized CSR in the form of a high-level law (Li & Appiah-Konadu, 2021). In 2008, the State-owned Assets Supervision and Administration Commission (SASAC) issued the CSR guidelines for state-owned enterprises directly under the central government mandating them to fulfill their corporate social responsibilities in fields such as corruption, environmental protection, and employee rights and to report on their CSR activities (Wu & Hąbek, 2021). In 2015, the Environmental Protection Law came into force to reduce the negative environmental impacts of business, which requires listed companies to publicly disclose their environmental protection activities (Khalid et al., 2022b). In 2018, the China Securities Regulatory Commission issued the revised Code of Corporate Governance for Listed Companies which formally established a framework for Environmental, Social, and Governance (ESG) disclosure (Wang, 2022). Finally, in 2022, the Measures for Enterprises to Disclose Environmental Information released by the Ministry of Ecology and Environment went into effect requiring enterprises to disclose environmental information (Zhang, 2022).
In general, corporations fulfill their social responsibility through internal and external CSR activities (De Roeck et al., 2014;Kim et al., 2021;Ye & Li, 2021). External CSR aims to improve the welfare of stakeholders outside the organization, such as consumers, the local community, the environment, and society at large (Brammer et al., 2007). In contrast, internal CSR activities focus on the well-being of internal stakeholders, in particular, employees, and include different instruments and policies in fields such as health and safety, diversity, sports and leisure, training and education, and work-life balance (De Roeck et al., 2014;Hameed et al., 2016). We focus on internal CSR as this domain is particularly relevant for the well-being of employees due to its potential to satisfy their basic needs, including safety and belongingness (Bauman & Skitka, 2012;Grabner-Kräuter et al., 2021).
Given that the intellectual roots of CSR lie in the Western world (Lorenzo-Molo & Udani, 2013), the literature highlights that CSR practices and stakeholder expectations in China may differ (Yin & Zhang, 2012). While in developed countries, the institutional setting typically ensures the protection of fundamental labor and social standards, the physical and psychological safety of employees in China cannot be taken for granted (Liu & Luo, 2021). This especially applies to blue-collar (rural migrant) workers in the manufacturing industry who are-due to poor education, low status, and legal discrimination-particularly vulnerable to adverse working conditions and exploitation by their employers (An & Bramble, 2018;Fitzgerald et al., 2013). Accordingly, mainstream Western-based internal CSR practices in terms of positive discretionary contributions to employee well-being (e.g., job enrichment, diversity or work-life-balance programs) are less relevant from the perspective of Chinese blue-collar workers than measures that are explicitly targeted at ensuring employees' physical and psychological safety (Lin-Hi et al., 2019).

Theoretical background and hypotheses
To derive our hypotheses, we first draw on social exchange theory (Blau, 1964) to elaborate on the direct link between internal CSR and work-related attitudes (H1a-c) and behaviors (H2a-c). Subsequently, we investigate the moderating effect of PRE-SOR importance (H3a-c and H4a-c) on these relationships based on the theory of PO fit (Kristof, 1996).
Research on social exchange theory (Blau, 1964) suggests that social exchange is a fundamental basis of human interactions. Based on the norm of reciprocity, individuals tend to return favors and resources and repay a favorable treatment by their organizations in a quid-pro-quo manner (Cropanzano & Mitchell, 2005;Ding & Chang, 2020). In general, the norm of reciprocity is a universal principle that is pervasive across cultures (Luo, 2002;Nguyen et al., 2022) and hence, is also relevant in the Chinese context (Li et al., 2014;Wu et al., 2006). Indeed, "[r]eciprocity norms are highly developed in China" (Starr et al., 2020, p. 69) and, given that members of collectivistic cultures tend to take reciprocity norms more seriously than members of individualistic cultures, it can be expected that Chinese employees are especially likely to reciprocate to their employers (Ng & Feldman, 2015;. Numerous investigations have successfully employed social exchange theory to predict the formation of Chinese employees' attitudes and behaviors Wang et al., 2018a;Zhang et al., 2022a).
In the workplace, social exchange relationships take the shape of an exchange of material and immaterial resources between employers and employees. These relationships evolve over time, when companies promote the well-being of their employees, therewith inducing them to react to the favorable treatment in a do ut des manner in terms of favorable work-related attitudes and behaviors (Cropanzano & Mitchell, 2005;Eisenberger et al., 2001). As outlined above, internal CSR is a way for employers to contribute to their employees' well-being. In consequence, internal CSR promotes social exchange processes and thus, should instill reciprocal behavior.

Job satisfaction
Internal CSR has the potential to satisfy some basic needs of employees, such as belongingness, security, and safety (Bauman & Skitka, 2012;Rupp et al., 2006). Therefore, CSR activities stimulate employees' feeling that their company cares about their well-being. In consequence, internal CSR should strengthen workers' felt obligation to reciprocate favorable treatment and enhance job satisfaction. Because blue-collar workers are typically exposed to very high levels of insecurity, internal CSR might be central to satisfy workers' safety and security needs and thus, lead to higher levels of job satisfaction: H1a Employees' perceptions of internal CSR are positively related to job satisfaction.

Trust in management
Trust in management is one of the primary outcomes of favorable exchange relationships between employers and employees (Renzl, 2008;Whitener et al., 1998). This is because reciprocity norms form the basis for trust building processes (Farooq et al., 2014;Kashyap & Rangnekar, 2016). Managerial trustworthiness is based on employees' perceptions of managers' ability (belief that management has influence, competence, and skills), benevolence (belief that management is caring and interested in doing good), and integrity (belief that management adheres to acceptable principles and upright values) (Mayer & Davis, 1999). As outlined above, internal CSR signals that the organization is benevolent, caring, and upright toward its employees, which conveys important information to employees not only about their current but also their future treatment. Accordingly, employees' positive expectations about their future well-being should increase their trust in the organization's main representatives (Dirks & Ferrin, 2002;Mayer & Davis, 1999). Hence, it can be expected that for blue-collar workers in China, who are faced with high levels of job insecurity and poor safety standards, internal CSR is an effective tool to enhance their perceptions of managerial trustworthiness: H1b Employees' perceptions of internal CSR are positively related to trust in management.

Perceived supervisory support
Employees' perceptions of the support from their direct supervisors in terms of their concern for the workers' well-being, i.e. supervisory support (Kottke & Sharafinski, 1988; are also relevant in this context. First, it seems plausible to assume that internal CSR affects the actual behavior of supervisors by making them more likely to care about the well-being of their subordinates. The existence of such a trickle-down effect has been outlined, for example, in the context of ethical leadership (Mayer et al., 2009). Second, since supervisors act as agents of the organization (Levinson, 1965), employees' perceptions of supervisory support are not only affected by the actual behavior of their supervisors but also by the behavior of their organization as such . In consequence, employees' favorable perceptions of their organization's supportive orientation may also instill favorable perceptions of their supervisors. This should particularly hold true in the Chinese context where workers' relationships with their immediate supervisors are the primary source through which they experience and define their relationship with their employers (Hui et al., 2004a). In sum, it can be expected that internal CSR enhances employees' perceptions of supervisor support by increasing both the actual level of supervisor support and employees' perceptions thereof: H1c Employees' perceptions of internal CSR are positively related to perceptions of supportive supervision.

Task performance
In line with the social exchange literature and the arguments above, we expect that internal CSR activities lead to an increase in employees' task performance because they signal that an organization fulfills its obligations toward its employees and supports them (Glavas & Kelley, 2014). In turn, employees feel obliged to reciprocate perceived support by enhancing their task performance .
We argue that the actual behavior of employers in the form of implemented and communicated internal CSR practices enhances workers' objective task performance. This is particularly likely to apply for workers in China who, due to the prevalence of reciprocity norms in China, are especially likely to pay back the favorable treatment by their employers in terms of enhanced task performance: H2a Internal CSR positively affects workers' task performance.

Voice behavior
We further look at extra-role behavior, i.e. employee behavior that goes beyond formalized role requirements and is geared toward promoting organizational objectives . Employees are more inclined to return favorable organizational treatment and go "the extra mile" when employers enhance their well-being via activities that exceed their expectations (Hui et al., 2004b). We focus on voice behavior, a form of extra-role behavior through which employees make constructive suggestions for improvements Liu et al., 2020) which is central in China (Farh et al., 2004).
In line with the reasoning above, internal CSR should increase workers' felt obligation and their willingness to make extra-effort to support organizational goals through constructive suggestions. Yet, employees are often reluctant to speak up because they are afraid that their voice behavior may be negatively perceived within the organization and undermine their internal status (Milliken et al., 2003). As outlined above, internal CSR fulfills employees' needs for safety and in doing so, contributes to a working environment in which employees feel safe to express their true opinion (Glavas, 2016). Hence, perceptions of psychological safety and the perceived need to reciprocate favorable treatment may increase employees' willingness to engage in voice behavior (Liang et al., 2012): H2b Internal CSR positively affects workers' voice behavior.

Complaints
As an important extension to the previous literature, we also look at complaints as a form of criticism behavior, that is fundamentally distinct from voice behavior . Complaints comprise behaviors that express employees' dissatisfaction with particular organizational practices and are used by employees to vent their emotions or achieve intrapsychic and interpersonal goals (Kowalski, 1996). Accordingly, in contrast to voice behavior, complaints are not aimed at initiating improvements and do not include any productive suggestions (LePine & Van Dyne, 1998).
As outlined above, internal CSR may lead to higher employee satisfaction as it fulfills employees' basic needs of safety and belongingness and strengthens their felt obligations toward their organizations. In contrast, a lack of internal CSR and the associated unsafe and unhealthy working conditions may lead to dissatisfaction and negative emotions that, in turn, result in more complaints (Kowalski, 1996): H2c Internal CSR negatively affects workers' complaint behavior.

The moderating effect of PRESOR importance
PO fit theory postulates that the congruency between the values of employees and the values of their organizations strengthens employees' satisfaction with organizational decisions (e.g., CSR activities), forms positive job attitudes, and leads to productive behavior (Chatman, 1989;Kristof, 1996). Meta-analytic findings support that PO fit improves employees' attitudes and behaviors (Kristof-Brown et al., 2005).
To account for PO fit, we focus on individuals' PRESOR importance in business (Etheredge, 1999). From the perspective of employees, internal CSR represents a signal about an employer's general attitude towards ethics and social responsibility. In line with PO fit theory, internal CSR should therefore be more relevant to and appreciated by employees high on PRESOR importance because they believe that CSR is the morally "right" action conducted by organizations. This fit between employees' PRESOR importance and their employers' internal CSR engagement thus can be expected to elevate the increase in job satisfaction, trust in management, perceived supervisory support, task performance and voice behavior. In contrast, for workers high on PRESOR importance, a lack of internal CSR might provoke even more complaints: H3 PRESOR importance strengthens the positive relationships between perceived internal CSR and (a) job satisfaction, (b) trust in management, and (c) supportive supervision.
H4 PRESOR importance strengthens the positive relationships between internal CSR and (a) workers' task performance and (b) workers' voice behavior, and the negative relationship between internal CSR and (c) workers' complaint behavior.

Study 1
Study 1 was designed to test the relationship between internal CSR and work-related attitudes (H1a-c), and the moderating effect of individuals' PRESOR importance (H3a-c). The proposed relationships were empirically tested based on data from a cross-sectional survey.

Sample and procedure
Data were collected from production workers in a large garment factory located in China. Participation was voluntary and took place outside of regular working hours, but participants were paid for an extra half an hour. In total, 930 blue-collar workers participated. Missing data for the study variables ranged between 2.5 and 9.6% for items from attitude measures, but was higher for demographic variables (maximum of 24.8% for age). To avoid a loss of statistical power inherent in the common procedure of listwise deletion (Graham, 2009;Roth, 1994), we included all participants with less than 10% missing values, resulting in a sample size of N = 803. Subsequently, we imputed missing data, using the mice package (Van Buuren & Groothuis-Oudshoorn, 2010) in R for multiple imputation (m = 5). Hypothesis tests were performed on each of the five datasets and the results were combined following the rules suggested by Rubin (1987).

Measures
We used 7-point Likert-scales for all survey items (unless indicated otherwise).

Perceived internal CSR
In order to measure perceptions of internal CSR, we constructed a scale with three items derived from Turker (2009) and two self-developed items. The scale reflects how responsibly the organization acts toward its employees in the eyes of individual employees. Due to workers' low education, we put particular emphasis on formulating the items as comprehensively as possible (see Appendix A for the wording of the items). The scale had a high internal consistency (Cronbach's α = 0.83).

PRESOR importance
We used the five items of the PRESOR subscale "importance of ethics and social responsibility" from Etheredge (1999) that accounts for specifics in the Asia-Pacific region. PRESOR importance in our study had a high internal consistency (Cronbach's α = 0.80).

Job satisfaction
To measure job satisfaction, we drew upon Zhou and George (2001) three-items scale. This job satisfaction scale with one reverse-coded item yielded a relatively low internal consistency (Cronbach's α = 0.56) in the context of Chinese blue-collar workers, which is substantially lower than Cronbach's α = 0.86 in the original study by Zhou and George's (2001). We address this issue below.

Supportive supervision
We used eight items to evaluate the degree to which workers felt supported by their supervisors. The items were adapted from Oldham and Cummings (1996). Internal consistency for this scale was high (Cronbach's α = 0.85).

Control variables
In line with previous research (Farooq et al., 2014;Valentine & Fleischman, 2008), we covered a wide range of individual differences and work-related factors to account for their potential impact on our outcome variables. Specifically, we controlled for age, gender (dummy with 1 for male), family status (dummy with 1 for being married), children (dummy with "1" for having children), ethnicity (dummy with "1" for Han nationality), education (six categories ranging from 1 = "no education" to 6 = "University degree"), organizational tenure, number of previous employers, and years of work experience. Table 1 shows the means, standard deviations, and zero-order correlations between all variables. Internal CSR is significantly and positively related with all three investigated work-related attitudes (ranging from 0.36 to 0.44).

Results
For hypothesis testing, we applied multiple linear regressions based on an Ordinary Least Squares (OLS) estimator, using the m = 5 data sets from multiple imputation. The results are displayed in Table 2. In Models 1a-c, we only included control variables. The total variance explained is about 2% for all three models and there is no consistent pattern of significant control variables. In Models 2a-b, internal CSR and PRE-SOR importance were added. Results indicate that internal CSR had a significantly positive relationship with job satisfaction (b = 0.053, p < 0.001), trust in management (b = 0.063, p < 0.001), and supportive supervision (b = 0.078, p < 0.001). This supports Hypotheses 1a-c. In Models 3a-c, the interaction term of internal CSR with PRESOR importance was added. As can be seen in Table 2, PRESOR importance strengthens the link between perceived internal CSR and job satisfaction (b = 0.003, p < 0.01), trust in management (b = 0.002, p < 0.01), and perceived supportive supervision (b = 0.003, p < 0.01). Thus, we find support for Hypotheses 3a-3c.
Despite having used a standard measure of job satisfaction, we obtained a relatively low internal consistency. To test for the robustness of the results, we ran additional analyses with a single-item measure of job satisfaction (Wanous et al., 1997). When this measure was employed, the results remained consistent with our multi-item measure of job satisfaction.

Study 2
Study 2 was designed to test the impact of internal CSR on work-related behaviors (H2a-c) and the moderating role of PRESOR importance (H4a-c) in an experimental setting.

Sample and procedure
The field experiment was conducted in the same garment factory based on a similar recruitment procedure as in Study 1. All employees participated voluntarily in the experiment in their off-work time and received overtime payment. In total, we acquired 160 participants who were randomly assigned to either the experimental group (n = 68) or the control group (n = 92). The experiment was conducted in seven small groups (three experimental and four control groups) over five days.
In the treatment group, we manipulated the presentation of the firm's internal CSR activities. Therefore, an interactive presentation was held by a Chinese employee from the factory's HR department who was trained by one of the authors before. The presentation covered descriptions of a broad range of internal CSR activities that the factory had implemented, including occupational health and safety measures, social audits, provision of additional non-financial benefits, leisure activities, and a grievance mechanism. In addition, workers were informed that the factory had a strong reputation for internal CSR, evidenced by evaluations from external experts and media reports. This was supported by a short video from an external expert who had personally visited the factory, had spoken with employees, and gave a very positive evaluation of the factory's internal CSR activities. The control group did not receive any information on the factory's CSR activities. Subsequently, all participants worked on two performance tasks (described below) and filled out a questionnaire.

Manipulation check
We used the same measure of perceived internal CSR as in Study 1. The five items were rated on a scale from 1 (strongly disagree) to 7 (strongly agree). As in Study 1, this scale yielded a high internal reliability (Cronbach's α = 0.83).

Task performance
We designed a "stamp task" to assess blue-collar workers' task performance, which we adapted from previous studies (Gerhards & Siemer, 2016;Gill et al., 2013). Participants received sheets of paper and a stamp with the company logo. Each sheet of paper contained 24 boxes and each box consisted of an outer frame and an inner frame (see Appendix B). Participants were informed that this stamp task is used by researchers to assess employee productivity in organizations. Participants were also notified that their employer was highly interested in a good task performance for reputational reasons. Accordingly, the participants were asked to complete as many stamp tasks as possible. We further emphasized that the stamp can only be counted if the company logo fell into the inner box on the sheet. Hence, it was possible to measure workers' task performance in two ways: (1) Stamp quantity, captured by the total number of stamps, and (2) Stamp quality which was assessed by the ratio between the number of valid stamps and total stamps.

Voice behavior and complaints
Participants were asked to voluntarily write down comments and suggestions for improvements in an open section at the end of the experiment. We developed a coding scheme based on the operationalization of voice and complaints recommended in prior research (LePine & Van Dyne, 2001). Specifically, voice behavior includes changeoriented and constructive suggestions which aim to improve an organization's policy, procedures, and actions (LePine & Van Dyne, 2001). Thus, voice behavior is mostly expressed with words such as "wish", "hope", or "suggest." An example of voice behavior included: "I suggest that the cafeteria should distribute the food as a whole set menu to workers because there might be safety issues when workers are trying to get each individual course themselves within a big crowd." Complaints, in contrast, reflect simple expressions of dissatisfaction or discontent (Kowalski, 1996) and do not include any change-related suggestions. A complaint example is: "The plates used in the cafeteria are always dirty." Two Chinese researchers coded the voice comments and complaints as numeric continuous variables based on the comment sections independently. Inter-rater reliability was high (interclass correlation coefficient ICC(2,2) voices = 0.86; ICC(2,2) complaints = 0.98). We used the average score from the two raters as measures of voice and complaints.

Control variables
The randomized allocation of individuals in our experiment reduces the utility of statistical controls (Becker, 2005). However, since H4a-c assume interactions with PRESOR importance, which is not controlled by experimental designs, we included gender, education, and organizational tenure as controls (with the same measures as in Study 1). Table 3 presents the means, standard deviations, and zero-order correlations of all study variables.

Results
Results from independent-samples t tests comparing group means and corresponding mean difference intervals based on bootstrapping of 1000 samples are displayed in Table 4. As can be seen, perceived CSR is significantly higher in the experimental group than in the control group, indicating a successful manipulation. The results further show that there are significant group differences for stamp quantity (M Control = 39.72, M Experimental = 46.01; p < 0.05) and stamp quality (M Control = 0.69, M Experimental = 0.57; p < 0.01). Hypothesis 2a states a positive impact of internal CSR on workers' task performance. While the results for stamp quantity support Hypothesis 2a, the effect for stamp quality goes in the opposite direction, i.e. lower stamp quality in the experimental group. Hypothesis 2b did not find support in the data as no significant mean differences were found for voice behavior (M Control = 0.17, M Experimental = 0.10; p > 0.10). Lastly, complaints were significantly higher in the control group (M Control = 0.51, M Experimental = 0.24; p < 0.05), supporting Hypothesis 2c.
In Hypotheses 4a-c, we assumed moderation effects of PRESOR importance on the relationships between internal CSR and work behaviors. Table 5 displays multiple linear regressions results. The interaction effect was only significant for voice behavior (Hypothesis 4c; b = 0.09, p < 0.05), but not for stamp quantity, stamp quality, and complaints.

Discussion and conclusion
We conducted a survey and a field experiment in a Chinese garment factory to investigate the effects of internal CSR on blue-collar workers in China. In the survey study, we found that workers with higher perceptions of internal CSR were more satisfied with their jobs, had more trust in management, and reported to receive more support from their supervisors. The field experiment revealed that internal CSR information led workers to generate a higher output and to complain less. However, internal CSR information also decreased workers' performance quality and did not yield any differences in workers' voice behavior. The results of the two studies further demonstrated that workers' attitudes toward the importance of ethics and social responsibility strengthened the positive relationship between internal CSR and job satisfaction, trust in management, and perceived supervisory support. In terms of work behaviors, PRESOR importance was found to significantly affect voice behavior in the way that workers, who perceived ethics and CSR to be important, engaged in more voice behavior and offered more constructive suggestions when exposed to internal CSR activities.

Theoretical contribution
Our work makes a number of important contributions to the literature. First, despite the fact that CSR has been constantly gaining importance in China, little is known about its benefits for Chinese business organizations. Thus, the findings of the present paper contribute to our knowledge about the individual-level outcomes of internal CSR in China. In line with the few existing findings of positive CSR outcomes in Table 3 Means, standard deviations, and correlations of all variables in Study 2
Second, previous findings have been, to a large extent, limited to cross-sectional survey studies which are subject to common method biases and cannot prove causal relationships between CSR and employee reactions (Jones et al., 2019). For instance, measures of employees' CSR perceptions might also be affected by the outcomes measured since employees who have more positive attitudes are also more likely to evaluate their employer more positively. We address this limitation in that our findings are partially grounded in data collected from a field experiment. Our experimental study allowed us to examine the direct causal effects of actual internal CSR instead of perceived CSR based on a natural intervention. A further strength of this study is that we have measured objective behavioral outcomes (performance, actual voice behavior, and complaints) compared to other CSR studies that have used subjective measures. For example, it is common practice to use questionnaires to measure task performance in terms of supervisors' ratings of their employees (e.g., John et al., 2019;Shao et al., 2019;Shen & Benson, 2016) and measure voice behavior via employee self-assessment (e.g., Hu & Jiang, 2018;Ilkhanizadeh & Karatepe, 2017;Wang et al., 2020a). Third, our findings extend what we already know about how employees respond to CSR. While our findings on job satisfaction and job performance quantity provide an important replication of existing findings (e.g., De Roeck et al., 2014;Shen & Benson, 2016) in a different context (here: Chinese blue-collar workers) and with a different method (here: experimental design), we further identify relevant outcomes that have not been studied before. Whereas previous studies have looked at organizational trust (Farooq et al., 2014;Silva et al., 2022), we show that trust in management and perceived supervisory support are also affected by internal CSR. This implies that internal CSR is also beneficial for leaders at different hierarchy levels who have not been considered sufficiently in theoretical and empirical CSR studies yet.
In addition, some of our results contradict previous empirical findings and theoretical frameworks. Interestingly, we found a negative relationship between internal CSR and performance quality. This counter-intuitive finding might be the result of a trade-off between speed and accuracy of the performance task in our field experiment (Howell & Kreidler, 1963). Considering the time pressure workers might have felt in our setting, it is possible that they favored a faster reaction time over result accuracy with the intention to contribute to overall firm success. It seems also reasonable that internal CSR exhibits effects similar to other extrinsic rewards, such as financial incentives, that were consistently shown to be related to performance quantity, but not necessarily to performance quality (Jenkins et al., 1998). As such, our finding emphasizes that the type of performance studied is relevant when analyzing whether CSR has positive or negative effects on employee performance. This underscores the need for further empirical investigations.
Furthermore, our study adds to previous research by uncovering additional internal CSR-related employee outcomes by showing that workers in the CSR treatment group were less likely to voice both their suggestions (yet insignificant) and grievances (significant) than workers in the control group. This might be attributed to the specifics of the Chinese culture. Compared to Western cultures, voice behavior in China may imply higher social costs due to the overarching importance of values such as harmony and solidarity in the relationship between workers and their employers (Liu et al., 2010;Wong, 2012). It might have been the case that workers feared that even making positive suggestions could be interpreted as a sign of criticism. Based on the assumption that employees value internal CSR, the CSR treatment in our experiment might have restrained employees from raising their voice. Another potential explanation is that Chinese blue-collar workers have the concern that they might lose their job quickly if they speak up, so that they might perceive the voicing of criticism as especially risky and costly and abstain from it altogether (An & Bramble, 2018;Qian et al., 2022;Yuan et al., 2021). Because empirical research on these relationships is lacking, future work is needed to corroborate whether internal CSR lowers voice behavior and complaints in other settings. We further extend previous research by showing that a fit between employers' and employees' attitudes regarding ethics and CSR matters. This underlines the importance of taking individual differences into account when investigating the effects of internal CSR on employee attitudes and behaviors and corroborates previous calls for more research on individual values in order to enhance our understanding of when and how CSR leads to beneficial effects for organizations (Jones et al., 2019).

Practical implications
Our study has also important practical implications given that increasing societal expectations and pressures regarding internal CSR are becoming an important factor for Chinese firms to secure their social license to operate, especially in relationship with governmental institutions (Li & Lu, 2020). Our findings indicate that enhancing the well-being of workers could be a sound managerial approach to improve work-related outcomes in the Chinese blue-collar sector. Therefore, we advocate for a stronger incorporation of CSR into the HR strategy of Chinese firms.
Furthermore, internal CSR can be a fruitful investment for Chinese factories to tackle some of their current challenges. In particular, Chinese factories are increasingly struggling to recruit and retain workers (Huang, 2011;Zhang et al., 2021c). In this respect, our findings suggest that internal CSR is an effective approach to respond to the shortage of blue-collar workers in Chinese firms. This is because positive attitudes, such as job satisfaction and trust in management, have the potential to reduce turnover intentions (Dirks & Ferrin, 2002;Griffeth et al., 2000). Given that managers in Chinese factories tend to neglect the well-being of their employees and often violate workers' rights (Yang, 2020), practicing internal CSR can be seen as an innovative approach that can help Chinese factories to generate a competitive advantage.
Based on the finding that employee PRESOR importance positively moderates the link between CSR and work outcomes, we expect that CSR will gain more relevance for Chinese HR management in the future. This is because the younger, better educated generation of Chinese employees has higher job expectations in terms of extrinsic and intrinsic work values (Huang et al., 2017;To & Tam, 2016;Zhang et al., 2021a) and in consequence, might attach greater importance to ethics and CSR performance of their employers. In light of the growing labor shortage in the blue-collar domain 1 and the associated rise in Chinese workers' bargaining power, CSR may allow Chinese businesses to better meet the changing demands of their workers in the future. Thus, internal CSR can be seen as an opportunity to adapt to ongoing changes in the Chinese labor market (Lin-Hi et al., 2019).
Lastly, the findings from our field experiment also highlight the benefits of communicating internal CSR activities to employees. Since blue-collar workers have usually little information about their company, the internal communication of what the company is doing with regard to internal CSR is of particular relevance. We show that informing employees about the benefits they have via internal CSR practices can be an effective HR tool for Chinese companies in order to increase working speed and lower complaints of employees.

Limitations and future research
As with every research, this work has limitations which, at the same time, offer fruitful avenues for future research. First, data was collected in only one factory which limits the robustness and generalizability of our findings. However, obtaining access to reliable data in China is difficult in general and in Chinese factories in particular (To & Tam, 2016). In light of the restricted access to data from Chinese factories, the dataset for this study can be considered as unique not only with regard to its nature but also because it was collected under the adherence to strict scientific rules, such as randomization, assurance of anonymity, and non-interference of the employer. Nonetheless, we acknowledge that a more representative sample would be beneficial for the verification of our results. We encourage scholars to investigate the effects of internal CSR on blue-collar workers in China.
Second, our results of Study 1 are, as other cross-sectional studies, limited in making causal claims. In our study, higher levels of job satisfaction, trust in management, and supervisory support might have led to more positive evaluations of internal CSR. However, we find similar results for our hypotheses on the moderating effects, which should not be subject to common method biases (Siemsen et al., 2010). Moreover, the results of our field experiment (Study 2) further support internal CSR's positive effects on workers' reactions.
Third, the performance measure in our stamp task was not directly related to workers' daily job tasks, i.e. sewing clothes. However, it was not possible to measure outcomes of daily job tasks without disturbing the production process and endangering anonymity. Hence, further research is needed to analyze the effects of internal CSR on job performance.
Fourth, the paper provides support for the relevance of internal CSR for Chinese blue-collar workers. However, although we elaborate on the relevance of social exchange mechanisms in the Chinese context, the theoretical framework of the study is still rooted in Western theories which are not per se applicable to other cultures. Thus, in order to fully understand the relevance of CSR in the Chinese business world, there is a need to develop theories that explicitly take into account the countries' specifics (Filatotchev et al., 2021;Yin & Quazi, 2018).
Fifth, since China is a culturally and economically highly heterogeneous country (Kwon, 2012;Xu et al., 2019), the relationships between internal CSR and work-related attitudes and behaviors might differ across regions. Thus, we encourage future research to empirically assess and compare the outcomes of internal CSR in different regions in and outside of China.

Performance measurement
Valid stamps: stamps for which the company logo fell into the inner frame of the stamp box Performance quantity: the total number of stamps Performance quality: the total number of valid stamps / the total number of stamps Funding Open Access funding enabled and organized by Projekt DEAL. The travel expenses of Nick Lin-Hi and Xiaohan Gao-Urhahn were covered by the company the study was conducted at. This company also paid the participants in the studies in the form of overtime payment (half an hour).