Perception of Corporate Social Responsibility, Organizational Commitment and Employee Innovation Behavior: A Survey from Chinese AI Enterprises

Abstract

This study delves into the relationships between the perception of corporate social responsibility (PCSR), organizational commitment and employee innovation behavior, as well as the multiple mediating roles of affective, normative and continuance commitment in the relationship between the perception of CSR and innovation behavior. This research involved 419 employees from 15 artificial intelligence (AI) enterprises in Shenzhen, China. This study’s hypotheses were tested using structural equation modeling. The findings indicate that PCSR significantly impacts innovation behavior, and affective, continuance and normative commitments also positively influence innovation behavior. Moreover, these three commitments play a partial mediating role in the relationship between PCSR and innovation behavior. This study enriches and expands the understanding of the multiple mediating mechanisms between PCSR and employee innovation behavior, providing a theoretical basis and guidance for management to comprehensively understand the role of employees’ PCSR in enhancing organizational commitment and fostering innovation behavior.

1. Introduction

Organizations face unprecedented opportunities and challenges in today’s era of rapid technological advancement (You et al. 2022; Flammer and Kacperczyk 2019). The significance of innovation is increasingly highlighted in organizations’ continual pursuit of sustainable competitive advantage (Li et al. 2023). The main asset of an organization is its employees (Cachón-Rodríguez et al. 2022). Innovation must be realized through innovative behavior. In today’s dynamic and cutthroat corporate circumstances, employee innovation behavior is essential to organizational success (Pujianto and Musyaffaah 2023).

Furthermore, as an integral part of the social system (Fernando et al. 2019), businesses must participate in socially conscious endeavors to facilitate their expansion (Barauskaite and Streimikiene 2021). Businesses are increasingly aware that social responsibility is related to enterprises’ fundamental interests and long-term development, and they consciously assume corporate social responsibilities (hereafter referred to as CSR) (Li et al. 2020; Ait Sidhoum and Serra 2018).

Employees’ values have also changed significantly, with subjective needs becoming more diverse and complex. Employees no longer solely pursue economic benefits; their perception of CSR initiatives undertaken by their organizations has become a pivotal factor influencing their attitudes and behaviors (Orazalin 2020; Van Dick et al. 2020). Consequently, does this stimulating factor also positively affect employees’ innovative behaviors? Existing research indicates a direct relationship between employees’ perceptions of CSR (hereafter referred to as PCSR) and their innovation (Hur et al. 2018; Li et al. 2020). However, the internal mechanisms linking PCSR and employee innovation behavior require further exploration (Wu et al. 2022; Chen and Zhang 2023).

Organizational commitment is frequently employed as a crucial psychological link in research on organizational behavior and attitudes (Lu et al. 2020). It represents employees’ significant intrinsic psychological attitudes (Wu and Chen 2018; Ahmed and Tahir 2019) and is related to organizational outcomes. Some studies indicate that PCSR may influence employees’ organizational commitments (Kong et al. 2019; George et al. 2021; Tran et al. 2021; Li and Qu 2023; Silva et al. 2023), while other research indicates that organizational commitment plays a crucial role in employees’ innovative initiatives (Nazir et al. 2018; Fauziawati 2021; Sharma et al. 2021; Chen and Liu 2022; Li and Qu 2023; Chtioui et al. 2023; Chen and Zhang 2023). Furthermore, studies reveal that employees are more likely to believe that the organization will treat them fairly and justly if they perceive the organization as treating key stakeholders ethically (Kim and Lee 2022). Employees are more likely to develop positive organizational commitments (Stojanovic et al. 2020) and respond with positive work behaviors (Choi et al. 2018). Thus, a mediating link between employee innovation behavior, PCSR and organizational commitment may exist.

This study utilizes the stimulus–organism–response (S-O-R) theory to examine the relationships among PCSR, organizational commitment and employee innovation behavior. The S-O-R theory framework, originating in behavioral psychology, is extensively applied in organizational behavior research (Ahmed et al. 2020; Huang et al. 2022). The S-O-R framework can be used to analyze how environmental stimuli effectively influence individuals’ internal states, subsequently eliciting individual behaviors (Xu and Wang 2020). In the organizational context of the S-O-R theoretical framework, stimuli encompass all external factors that induce employee changes (Tang et al. 2019; Yin et al. 2021). PCSR is considered a type of stimulus (Wu et al. 2022; Chen and Zhang 2023; Nguyen-Viet et al. 2024). At the same time, the S-O-R framework further indicates that the organism typically acts as a mediator between stimulus and response (Bigne et al. 2020). Psychological states often represent the organism (Ahn and Seo 2018; Leung and Wen 2020; Leung et al. 2021). The response (R) represents the outcome, including various behavioral feedback responses (Chen and Zhang 2023). Following this logic, employees’ perceptions affect their behaviors through mediating factors, usually psychological attitudes (Wu et al. 2022; Chen and Zhang 2023). Organizational commitment is regarded as a psychological attitude of employees (Ahmed and Tahir 2019; Wu and Chen 2018; Chen and Zhang 2023). The concept of commitment is commonly divided into affective commitment, normative commitment, and continuance commitment (George et al. 2021). Some studies have also confirmed the efficacy of organizational commitment as a mediating variable (Fauziawati 2021; Chen and Liu 2022; Li and Qu 2023). Therefore, we predict that organizational commitment positively impacts employee innovation behavior and mediates between PCSR and employee innovation behavior. We utilize the S-O-R framework to analyze the connection between PCSR as a stimulus and employee innovation behavior as a response, along with the mediating role of organizational commitment (organism) in this relationship.

The survey population for this study consists of employees from AI enterprises in Shenzhen, China. Presently, China possesses the world’s second-largest cluster of artificial intelligence enterprises and has emerged as one of the global leaders in the AI industry (Dong et al. 2020). As a key city in China’s AI industry, Shenzhen has become a leader in setting domestic AI industry standards (Li 2024). As of April 2024, Shenzhen has led or participated in formulating six out of ten national standards published in the AI field and has ranked first in international patent applications among Chinese cities for 20 consecutive years (Li 2024; He 2024). Due to the need for economic transformation, China places a heightened emphasis on artificial intelligence. The AI industry relies on continuous innovation to survive in a competitive landscape (Dong et al. 2020). However, research analyzing the innovation development of Chinese AI enterprises is quite limited (Arenal et al. 2020), highlighting the necessity to address the gap in studies on employee innovation behavior in Chinese AI enterprises. Shenzhen, China’s next-generation AI innovation development experimental zone and a national AI innovation application pilot zone (Shenzhen Artificial Intelligence Industry Association 2024), provides an appropriate context to study the relationships among employees’ PCSR, organizational commitment and innovation behavior.

Therefore, this study uses multiple mediation analyses to examine the relationship and influence paths between PCSR and employee innovation behaviors in AI firms in Shenzhen. It also aims to confirm whether different types of organizational commitments mediate the relationship between PCSR and employee innovation behaviors. Although CSR has existed for over 60 years, research on CSR has grown significantly recently (Escamilla-Solano et al. 2023). The findings of this study expand research on CSR at the micro level. Although existing literature has utilized the S-O-R framework to examine the relationship between PCSR and employee behavior, it has primarily focused on organizational trust (Nguyen-Viet et al. 2024), organizational identification (Wu et al. 2022), work enthusiasm (Alhumoudi et al. 2023), and well-being (Su and Swanson 2019; Ahmed et al. 2020). The S-O-R framework is used in this study to show how PCSR can lead to innovation behavior through the important role of organizational commitment. It gives us a new way to think about how CSR affects employee behavior and adds to our understanding of how the different aspects of organizational commitment strengthen the effect of PCSR on innovation behavior. This study also adds to the S-O-R literature and expands the S-O-R framework. Additionally, by selecting the organizational context of Chinese AI enterprises, this study enriches and optimizes the sample pool for research on organizational commitment and employee innovation behavior within the Chinese context. It also provides more comprehensive insights into the interactions among Chinese employees’ PCSR, the dimensions of organizational commitment and innovation behavior.

2. Literature Review

2.1. Theoretical Framework

This study explores the relationships between employees’ PCSR, organizational commitment and innovation behavior based on the stimulus–organism–response (S-O-R) theoretical framework. Mehrabian and Russell (1974) proposed the S-O-R framework based on stimulus–response theory, positing that environmental stimuli influence an individual’s internal state, affecting their behavior. The S-O-R framework provides insights into how stimuli shape internal states in the human environment and influence responses represented by behavior (Chen and Zhang 2023). According to the S-O-R theory, changes in internal drives and mechanisms brought about by external stimuli determine human behavior (Pandita et al. 2021). When individuals receive stimuli, internal psychological processes are activated, leading to behavioral changes (Ferdous et al. 2021). Under certain stimulus conditions, a person typically generates corresponding subsequent behaviors mediated by an internal state (Ahmad et al. 2021). In other words, the organism often acts as an intermediary between stimulus and response (Bigne et al. 2020).

Studies utilizing the S-O-R framework model have increasingly enriched and been widely applied in organizational behavior research (Laesser et al. 2019; Ahmed et al. 2020; Huang et al. 2022). In the existing research, the PCSR has been confirmed as one of the external stimuli employees receive (Wu et al. 2022; Nguyen-Viet et al. 2024; Chen and Zhang 2023). The organism is considered an individual’s internal state after exposure to these stimuli, representing a psychological state or mechanism (Ahn and Seo 2018; Leung and Wen 2020; Leung et al. 2021). Therefore, in the relationship between PCSR and innovation behavior, PCSR can be regarded as a stimulus influencing employee behavior and response (Ahmad et al. 2021). As an important part of employees’ mental health, organizational commitment can be seen as an organism (Ahmed and Tahir 2019; Wu and Chen 2018; Lu et al. 2020; Chen and Zhang 2023) that serves as a mediator in these relationships linking the effects of PCSR (S) to employee innovative behavior (R). Therefore, this study considers the S-O-R framework as suitable for explaining the relationship between employees’ PCSR, organizational commitment and innovation behavior.

2.2. Perceptions of CSR and Employee Innovation Behavior

Corporate social responsibility (CSR) has been widely discussed in global business, but there is yet to be a universally accepted definition of CSR (Barauskaite and Streimikiene 2021). Carroll (1979) defines CSR as the societal expectations placed on organizations at a particular time regarding economic, legal, ethical and voluntary obligations. El Akremi et al. (2018) view CSR as targeted organizational activities and policies that aim to improve stakeholders’ well-being by considering the triple bottom line of economic, social and environmental performance. CSR guides and encourages companies to behave more responsibly, positively impacting stakeholders such as the environment, customers, employees and communities. By fulfilling social responsibilities, companies can achieve sustainability in their business strategies (Ahmed and Tahir 2019). Barauskaite and Streimikiene (2021) stated that CSR should be considered a broad process. Socially responsible businesses align their interests with those of shareholders, employees, consumers, the environment, communities and other stakeholders.

Research has shown that CSR can yield macro-level outcomes but can also be seen as a micro-level issue related to employee outcomes (Hur et al. 2018; Escamilla-Solano et al. 2023). When conducting micro-level studies of CSR, it is crucial to understand employees’ perceptions of corporate CSR behaviors. Employees’ cognitive view of CSR is commonly called their perception of CSR (PCSR) (Ng et al. 2019; Hur et al. 2018). Even though both objective CSR and PCSR are important, PCSR may have a bigger effect on how employees react than objective, real CSR actions that employees may need to be made aware of (Gond et al. 2017), which can have a significant effect on how employees feel, what they do and how well they do their job (Hur et al. 2018). PCSR, as a predictor of various work outcomes, consistently outperforms objective CSR (Ng et al. 2019). Therefore, this study conducts an analysis based on PCSR, which is PCSR as employees’ subjective feelings and opinions on the CSR measures implemented by the organization.

Employee innovation behavior, as a form of employee behavioral outcomes, has garnered widespread attention in recent years. Employee innovation behavior was defined by Scott and Bruce (1994) as the actions of employees in developing, launching or implementing innovative strategies, procedures, services or products at work. Kwon and Kim (2020) define innovation behavior as the deliberate proposal and application of new and better ideas, procedures, practices and policies to achieve long-term sustainability, corporate success and organizational effectiveness. Innovation behavior is also considered an additional role or discretionary action (Cho and Song 2021). Proposing and implementing new and better ideas, procedures, practices and policies for the organization’s efficacy, financial success and long-term sustainability is a conscious representation of employee innovation behavior (Nazir et al. 2018); it represents spontaneous actions to improve or create new situations.

Researchers have indicated that implementing CSR can alter employees’ attitudes and behaviors (Tong et al. 2019). CSR can enhance employees’ creativity by influencing their perceptions of the work environment (Kim and Lee 2022). Companies actively engaged in CSR typically advocate for a work atmosphere that is unrestricted and liberating (Rongbin et al. 2022). CSR activities of a company can also create an environment of confidence and trust for employees, enabling them to take risks without fearing negative consequences (Mao et al. 2021), thus making them more inclined to engage in innovative behaviors and produce more innovative products and services that benefit both the company and the community (Rongbin et al. 2022). CSR practices also communicate a positive corporate image and reputation to employees (Ali et al. 2020; Javed et al. 2020), which employees internalize as part of their values, making them more willing to embrace opportunities and challenges and thereby encouraging more innovative behaviors (De Roeck and Farooq 2018). Simultaneously, the implementation of CSR is an organizational learning process. In fulfilling their CSR, organizations must coordinate relationships among various stakeholders, foster a more open organizational culture, and promote communication and interaction internally and externally, which helps employees generate innovation behaviors (Li et al. 2020). Richter et al. (2021) argue that companies’ long-term CSR investment behaviors can stimulate the drive for product innovation, development, and differentiation.

Furthermore, CSR can increase the social acceptance of an organization, leading to higher-quality relationships within the institution and more being conducive to building social capital (Blanco-Gonzalez et al. 2020). Social capital is a driving factor in innovation development (Pylypenko et al. 2023). Through information exchange and technical cooperation, employees with social capital can affect team or organizational operations. They can also obtain more industry and market information, stimulating employees’ innovation behavior (Zhao et al. 2022). The following hypotheses are put forth in light of the study above:

H1. 

PCSR positively influences employee innovation behavior.

2.3. Perceptions of CSR and Organizational Commitment

Employees’ attachments to the organization and their jobs are characterized by a psychological condition called organizational commitment (Ahmed and Tahir 2019; Wu and Chen 2018). Organizational commitment was characterized by Mowday et al. (1979) as a strong desire to stay a member, a readiness to invest significant work for the organization, and a staunch belief and acceptance of the organization’s aims and ideals. Employee commitment encompasses employees’ desires, feelings, and beliefs about participating in the organization. It goes beyond mere alignment with goals and values; it also involves obligations, needs, and desires to demonstrate involvement with the company (Ahmed and Tahir 2019). Meyer and Allen (1991) postulated three categories of organizational commitment: affective commitment, normative commitment and continuance commitment. Affective commitment is the term used to describe a person’s identity, emotional attachment and participation in their employment. Normative commitment is the feeling of duty a person has to their organization. Continuance commitment indicates that employees believe they need to stay with the organization. Employees who demonstrate continuance commitment are loyal to their organization and willing to stay with it (Zaman and Nadeem 2019; George et al. 2021). These three commitments each have distinct assessment criteria and represent the various psychological states of particular employees (Ahmed and Tahir 2019).

Kim and Lee (2022) point out that CSR engagement conveys positive values. Employees are more committed to the organization when they actively believe in its principles, since they are reluctant to leave a place where they have a great work environment. Altruistic CSR attribution positively impacts employees’ attitudes. Employees will feel pleased to work for a business if various stakeholders value and acknowledge its CSR initiatives and social standing (Tian and Robertson 2019). When CSR projects or behaviors meet employees’ expectations, they can enhance individuals’ self-worth and satisfy their pursuit of positive self-esteem. They could be more committed to the organization because of their improved attitude (Khan et al. 2021). Employees’ favorable perceptions of organizational CSR ought to result in a favorable perception of organizational justice (George et al. 2020), thereby enhancing emotional connections to the company and deepening a sense of responsibility (Diatmono 2019), contributing to forming normative commitment.

Furthermore, investing resources in social and charitable activities or employee care can enhance employees’ happiness, positively impacting their commitment (Wu et al. 2019; Ahmed and Tahir 2019). According to Tran et al.’s (2021) research, employees’ affective commitments increase with their levels of empowerment, health and safety, job satisfaction and work–life balance. CSR also helps to build social capital (Blanco-Gonzalez et al. 2020). Social capital is a resource that retains members within an organization (Cachón-Rodríguez et al. 2022). According to Barauskaite and Streimikiene (2021), CSR can also support corporate sustainable development. Employees’ long-term loyalty to the company will increase when they accept socially responsible actions (Cachón-Rodríguez et al. 2022), strengthening their continuance commitment. The following hypotheses are put forth in light of the study above:

H2. 

PCSR positively influences affective commitment.

H3. 

PCSR positively influences normative commitment.

H4. 

PCSR positively influences continuance commitment.

2.4. Organizational Commitment and Employee Innovation Behavior

Positive behaviors are the outcomes of positive attitudes (Khaleel et al. 2017). Employees with positive psychological attitudes toward the organization actively engage in the company’s success and sustainability, leading to positive actions (Wu et al. 2022). Organizational commitment, in essence, reflects employees’ positive psychological attitudes towards the organization, indicating their willingness to participate and become part of the organization (Wu and Chen 2018; Ahmed and Tahir 2019; Lu et al. 2020). Highly effective commitment employees had a higher chance of long-term employment retention and good work behaviors (Ghasempour Ganji et al. 2021). Employees with a strong normative commitment also typically show greater satisfaction and loyalty (Cao 2023). Loyal employees in the company exhibit positive extra-role behaviors, including innovation behavior. They also tend to be good achievers, innovating, developing and satisfying customers’ expectations (Cho and Song 2021). The following hypotheses are put forth in light of the study above:

H5. 

Affective commitment positively influences employee innovation behavior.

H6. 

Normative commitment positively influences employee innovation behavior.

H7. 

Continuance commitment positively influences employee innovation behavior.

2.5. The Mediation Role of Organizational Commitment

The S-O-R theory explains the relationship between employees’ PCSR, organizational commitment and innovation behavior. The S-O-R theory has been widely applied in studies of employee behavior (Attiq et al. 2017). The S-O-R theory emphasizes that human behavioral responses are complex (Zhu et al. 2020). Individuals will exhibit corresponding follow-up behaviors in response to specific stimulus conditions, which require cognitive processes mediated by conduction (Bigne et al. 2020; Wu et al. 2022; Chen and Zhang 2023). In other words, the human behavioral process involves a sequence from stimulus to changes in psychological state and then to action response (Ferdous et al. 2021; Ahmad et al. 2021). Employees’ PCSR can be considered a stimulus factor (Wu et al. 2022; Nguyen-Viet et al. 2024; Chen and Zhang 2023), influencing both their organizational commitment (O) and their innovation behavior (R). According to Ahmed and Tahir (2019) and Lu et al. (2020), organizational commitment, which acts as an organism (O) in the S-O-R framework, describes how a person thinks about and reacts to their surroundings. It mediates the connection between PCSR (S) and employee innovation behavior (R). Many scholars in previous studies have confirmed the mediating role of organizational commitment as an intermediate variable (Fauziawati 2021; Chen and Liu 2022; Li and Qu 2023). Based on the above analysis and incorporating the three-factor structure theory of organizational commitment, the following hypotheses are put forth in this study:

H8. 

Affective commitment mediates the relationship between PCSR and employee innovation behavior.

H9. 

Normative commitment mediates the relationship between PCSR and employee innovation behavior.

H10. 

Continuance commitment mediates the relationship between PCSR and employee innovation behavior.

Based on this study’s theoretical framework and literature review, Figure 1 shows this study’s complete theoretical model and research hypotheses.

3. Methodology

3.1. Sample and Data Collection

This study’s sample data were collected from AI enterprise employees in Shenzhen, China. China is undergoing economic transformation, driving increased attention toward artificial intelligence (AI) due to its capability to overcome the physical limitations of capital and labor, thereby opening up new sources of value and growth (Arenal et al. 2020). Consequently, the AI industry is rapidly expanding in China. According to the “Artificial Intelligence Development White Paper” published by the Shenzhen Artificial Intelligence Industry Association (2024), China ranks second globally in the number of AI companies. Shenzhen is recognized as one of the key cities for AI development in China, serving as a leader and demonstration area in developing the Chinese AI industry (Shenzhen Artificial Intelligence Industry Association 2024). It has achieved significant results in AI-related policy-making on the industry scale, as well as technological innovation (Li 2024; He 2024). Innovation primarily drives AI enterprises (Dong et al. 2020). This study selects employees of AI companies in Shenzhen as the research subjects for employee innovation behavior, which is representative and can effectively achieve the research objectives.

This study utilized questionnaire surveys to collect data. At the end of 2023, AI companies in Shenzhen employed approximately 140,000 employees (Shenzhen Artificial Intelligence Industry Association 2024). Following the method used by Krejcie and Morgan (1970) to determine the minimum sample size, the calculated minimum sample size was 385 individuals. To account for potentially invalid data in the returned questionnaires, the study planned to distribute 450 questionnaires to ensure that the number of valid responses was equal to or greater than 385.

During the sampling process, according to the “Artificial Intelligence Security Standardization White Paper (2019 Edition)” issued by the China Information Security Standardization Technical Committee (2019), the artificial intelligence industry chain is primarily divided into three levels: the basic layer, the technical layer and the application layer. Therefore, we first stratified the AI enterprises in Shenzhen according to these three layers. Based on the “2024 Artificial Intelligence Development White Paper” and data from the Qichacha database, it was found that by the end of 2023, approximately 20% of Shenzhen’s AI companies were in the basic layer, 15.6% were in the technical layer and 64.6% were in the application layer. This distribution determined the number of questionnaires to be distributed for each layer: 90 for the basic layer, 70 for the technical layer, and 290 for the application layer, totaling 450. After data stratification, we extracted detailed information on enterprises in each layer from the Qichacha database. Then, we randomly sorted the enterprises and allocated corresponding number ranges based on the scale ratio of each company. Subsequently, we randomly selected these numbers and distributed questionnaires to the companies falling within the selected number ranges. Each number corresponded to one sampling opportunity, and for each random sampling, the sample unit consisted of 10 employees. Ultimately, we distributed 90 questionnaires to the basic layer, involving 4 enterprises; 70 questionnaires to the technical layer, involving 2 enterprises; and 290 questionnaires to the application layer, involving 9 enterprises. A total of 450 questionnaires were distributed to 15 enterprises.

Before the survey, the researcher explained the academic purpose to each participant and assured them that their answers would be kept confidential and used for academic purposes only. All the participants provided consent. There were two stages to the data-gathering process: the participants filled out Questionnaire A in the first phase, including data on PCSR and demographic factors. Questionnaire B on employee organizational commitment and innovation behavior was distributed to the participants one month later for the second stage. Before formally distributing the questionnaires, we conducted a small-scale offline pilot test. In the pilot test, we assessed the reliability and validity of the questionnaire and confirmed the stability and consistency of the scales. The questionnaire’s reliability and validity were validated by analyzing the pilot test with a small sample, consistent with the hypotheses. Therefore, we confirmed the applicability of the questionnaire and used it as the survey instrument for data collection. After excluding invalid questionnaires, we obtained a final count of 419 valid questionnaires, resulting in an overall response rate of 93%.

The demographic information of the participants is shown in

Table 1. Demographic variable data.

Variable	Category of Statistics	F	%
Gender	Male	222	53.00
	Female	197	47.00
Education	Diploma and below	60	14.30
	Bachelor’s degree	206	49.20
	Master’s degree	107	25.50
	Doctorate and above	46	11.00
Length of Service	3 years and below	146	34.84
	4–6 years	212	50.60
	7–9 years	55	13.13
	10 years and above	6	1.43
Age	30 and below	152	36.30
	31–40 years old	201	48.00
	41–50 years old	56	13.30
	50 above	10	2.40
Total		419	100

. Among the survey participants, there were 222 males and 197 females, with 80% of the respondents primarily below the age of 40 years. More than 50% of the employees had 4 to 6 years of work experience. Employees with a bachelor’s degree or higher accounted for approximately 85.7% of the sample. It can be observed that the gender ratio in the survey sample was balanced, with the majority of the respondents being young and middle-aged, having considerable tenure in the company, and possessing a deep understanding and experience of their work and the company. Furthermore, most of them had received higher education, indicating a certain understanding of the concepts involved in this study. They were the primary drivers of innovative organizational initiatives and demonstrated high CSR participation.

3.2. Variable Measurement

This study employed the traditional “translate–back-translate” method to convert the English survey questionnaire into Chinese, ensuring that the Chinese participants could understand the questionnaire items accurately. We used a Likert five-point scale (1 = “strongly disagree” to 5 = “strongly agree”) to obtain responses from the participants. The measure scales for each variable are described below.

3.2.1. Dependent Variable

The dependent variable in this study was employee innovation behavior (abbreviated as EB), measured using a six-item scale from Iqbal et al. (2020), derived from the classic scale by Scott and Bruce (1994). The sample items for this scale included “I will actively seek to apply new technologies, procedures, or methods”. For detailed scale item contents, please refer to Appendix A.

3.2.2. Independent Variable

The independent variable in this study was PCSR. The scale for measuring this variable was sourced from Tian and Robertson (2019) and derived from the classic scale by Turker (2009). This scale included 12 items, with sample items such as “Our company contributes to endeavors and projects that promote societal well-being”. For detailed scale item contents, please refer to Appendix A.

3.2.3. Mediating Variable

The mediating variables in this study were affective commitment (AC), normative commitment (NC) and continuance commitment (CC). The scales used to measure these variables originated from Loan (2020), based on the traditional three-factor organizational commitment scale developed by Meyer et al. (1993). Each commitment scale comprised six items, with sample items such as “I feel emotionally bonded with my company”, “I feel the obligation to remain with my current employer” and “Currently, staying with my company is a matter of necessity as much as desire”. For detailed scale item contents, please refer to Appendix A.

3.3. CMB Test

Using self-reported data may introduce CMB (Kim and Lee 2022). However, this study conducted the Harman single-factor test (Podsakoff and Organ 1986), revealing that the variance explained by the first factor prior to rotation was less than 50%. Additionally, the correlation coefficients among the constructs were all below 0.90, indicating that CMB is not significant in this study.

4. Results

4.1. Descriptive Statistics

This study conducted a descriptive analysis test. As shown in

Table 2. Descriptive analyses.

Variable	n	Mean Statistic	Std. Deviation Statistic	Tolerance	VIF
PCSR	419	4.046	0.037	0.761	1.315
AC	419	3.883	0.042	0.813	1.229
NC	419	3.697	0.040	0.885	1.130
CC	419	3.722	0.042	0.775	1.291
EB	419	3.622	0.046	-	-

, it is clear that all the variable mean values were greater than 3. Among these, employee innovation behavior had the lowest mean value, and PCSR had the highest, demonstrating a high degree of agreement and readiness to engage among the respondents. Most of the items’ standard deviations fell between 0.8 and 1, indicating a somewhat concentrated data distribution with a moderate dispersion. The tolerance values of the independent variables were all above 0.20, and the VIF values were below 10, indicating that no multicollinearity issues among the variables were found (Thompson et al. 2017).

4.2. Reliability and Validity Test

This study conducted a confirmatory factor analysis (CFA) before testing the hypotheses. The Cronbach’s alpha values for all the constructs in

Table 3. Convergent reliability and validity.

Constructs	Items	Factor Loading	AVE	CR	Cronbach’s Alpha
PCSR	PCSR1	0.766	0.504	0.924	0.923
	PCSR2	0.715
	PCSR3	0.688
	PCSR4	0.684
	PCSR5	0.692
	PCSR6	0.733
	PCSR7	0.724
	PCSR8	0.653
	PCSR9	0.678
	PCSR10	0.717
	PCSR11	0.772
	PCSR12	0.693
AC	AC1	0.771	0.555	0.882	0.881
	AC2	0.728
	AC3	0.759
	AC4	0.715
	AC5	0.774
	AC6	0.720
NC	NC1	0.713	0.502	0.858	0.858
	NC2	0.717
	NC3	0.705
	NC4	0.684
	NC5	0.706
	NC6	0.723
CC	CC1	0.719	0.516	0.865	0.865
	CC2	0.713
	CC3	0.693
	CC4	0.736
	CC5	0.743
	CC6	0.706
EB	EB1	0.739	0.547	0.879	0.884
	EB2	0.691
	EB3	0.778
	EB4	0.753
	EB5	0.711
	EB6	0.763

were higher than 0.70, the composite reliability (CR) value was higher than 0.8 and the average variance extracted (AVE) value was higher than 0.50. The measurement model is reliable and convergent (Fornell and Larcker 1981; Kim and Lee 2022). As shown in

Table 4. Discriminant validity.

Variable	PCSR	AC	NC	CC	EB
PCSR	0.710
AC	0.356 **	0.745
NC	0.241 **	0.269 **	0.708
CC	0.420 **	0.319 **	0.258 **	0.719
EB	0.504 **	0.470 **	0.448 **	0.476 **	0.740

** p < 0.01. Off-diagonal elements are the correlation estimates. The bold items on the diagonal represent the square root of AVE.

, the correlation coefficient between each variable was less than the square root of the AVE value. Therefore, the measurement model of this study had strong discriminant validity (Fornell and Larcker 1981). The questionnaire reached acceptable levels for the study in various reliability and validity indicators, indicating good reliability and validity. Additionally, the relationships among the variables were significantly positively correlated (Table 4), providing preliminary support for our hypotheses.

4.3. Structural Model Fit Test

This study used a structural equation model, which was tested using the maximum likelihood estimation method. The fit indices for the structural route model used in this study are shown in

Table 5. Results of model fit test.

χ2	df	χ2/df	CFI	TLI	IFI	GFI	SRMR	RMSEA
1023.673	587	1.7439	0.9376	0.9422	0.8651	0.8867	0.0553	0.0422

. All the fit indices combined were within the typical range. The model had excellent goodness of fit (Hu and Bentler 1998).

4.4. Path Analysis and Hypothesis Test

To validate these hypotheses, we first conducted a path analysis.

Table 6. Standardized path estimates and hypothetical results.

Path	Standardized Coefficient	Standard Error	T-Value (C.R.)	p-Value	Hypothesis Result
PCSR → EB	0.260	0.070	4.602	***	H1 accepted
PCSR → AC	0.407	0.066	7.155	***	H2 accepted
PCSR → NC	0.282	0.063	4.975	***	H3 accepted
PCSR → CC	0.474	0.069	7.944	***	H4 accepted
AC → EB	0.265	0.052	5.403	***	H5 accepted
NC → EB	0.309	0.055	6.314	***	H6 accepted
CC → EB	0.263	0.056	5.057	***	H7 accepted

*** p < 0.001.

and Figure 2 display the path coefficients and significance levels. The analysis results show that PCSR had a significant positive impact on EB, thus supporting H1. Additionally, PCSR had a significant positive impact on AC, NC and CC, with PCSR having the strongest impact on continuance commitment. These result supports hypotheses H2, H3 and H4. AC, NC and CC also had significant positive impacts on EB, with normative commitment having the strongest impact on employee innovation behavior, thus supporting hypotheses H5, H6 and H7.

4.5. Mediation Effect Analysis

This study used the Bootstrap method, which is widely acknowledged as an efficient technique for analyzing mediation effects. A total of 5000 bootstrap resamples were specifically run to extensively analyze the numerous mediation effects in the model. A 95% bias-corrected confidence interval was produced using technique (BC 95% CI). Comprehensive data on total, direct and indirect impacts are included in

Table 7. Total, direct and indirect effects.

Indirect Effect	Point Estimate	BC 95% CI		p-Value	Hypothesis Result
		Lower	Upper
PCSR → AC → EB	0.108	0.062	0.165	0.003	H8 accepted
PCSR → NC → EB	0.087	0.049	0.148	0.002	H9 accepted
PCSR → CC → EB	0.125	0.067	0.186	0.005	H10 accepted
Direct Effect
PCSR → EB	0.26	0.127	0.355	0.005
Total Effect
PCSR → AC → EB	0.368	0.246	0.467	0.004
PCSR → NC → EB	0.347	0.240	0.451	0.003
PCSR → CC → EB	0.385	0.272	0.469	0.006
R-square
AC				0.166
NC				0.080
CC				0.225
EB				0.536

.

The mediation analysis results indicate that all the paths of indirect effects were significant, along with significant direct effects. Therefore, affective, normative and continuance commitment partially mediate the relationship between PCSR and employee innovation behavior, confirming H8 to H10.

In summary, the results of this study demonstrate that the questionnaire was reliable and valid, and the model fit well. All the paths showed significant impact effects, indicating the presence of mediating effects. Therefore, all the hypotheses proposed in this study were validated.

5. Discussion

5.1. Main Findings and Theoretical Contributions

This study aimed to analyze how employees’ PCSR influences their innovation behavior and to explore the multiple mediating roles of organizational commitment between PCSR and employee innovation behavior. Specifically, this research focused on a micro-level study of CSR, where employees’ emotional responses and behaviors towards CSR are shaped by their cognitive evaluations rather than merely the objective level of CSR activities. Therefore, when approaching CSR from a micro perspective rather than a macro one, it is crucial to prioritize PCSR over objective CSR measures. This approach aligns with the views of scholars such as Kim and Lee (2022), Li et al. (2020) and Ng et al. (2019).

The research results first confirm that employees’ PCSR significantly influences their innovation behavior. The higher the employees’ PCSR, the more innovation behavior they exhibit at work. This finding aligns with the research of Hur et al. (2018) and Li et al. (2020). This study considers that CSR improves relationships with various stakeholders and increases internal and external social capital (Blanco-Gonzalez et al. 2020), which can enhance employee innovation behavior (Zhao et al. 2022). On the other hand, CSR activities elevate corporate reputation (Ali et al. 2020; Javed et al. 2020), influencing employees’ behaviors and values (De Roeck and Farooq 2018; Van Dick et al. 2020). CSR activities can also improve the work environment (Mao et al. 2021), making employees more willing to actively explore and apply new technologies, processes or methods at work (De Roeck and Farooq 2018; Rongbin et al. 2022), which helps stimulate innovation behavior. Moreover, CSR programs often require close communication and collaboration among various departments and employees, fostering teamwork and collaborative innovation within the company (Li et al. 2020) and further encouraging employee innovation behavior.

Secondly, the findings of this study also validate the positive impact of employees’ PCSR on organizational commitment. This study posits that when organizations actively undertake CSR, employees become more closely linked with the company principles as they see the benefits of these efforts (Tian and Robertson 2019). CSR fosters the alignment of employee values with the organization, creating a strong emotional connection that enhances affective and normative commitment (Diatmono 2019). At the same time, CSR also helps to build social capital and promote corporate sustainable development (Blanco-Gonzalez et al. 2020; Barauskaite and Streimikiene 2021). It helps retain internal members and enhance employee loyalty (Cachón-Rodríguez et al. 2022), which will help enhance employees’ continuance commitment. Moreover, this study found that the impact of PCSR on continuance commitment is the most significant, which further demonstrates that a positive CSR experience can enhance employees’ willingness to continue serving the company.

Thirdly, this study confirms the positive impact of organizational commitment on employee innovation behavior, with normative commitment having a more significant influence on employee innovation behavior, reflecting that employees with a sense of obligation are more willing to innovate. Additionally, this study reveals the mediating roles of organizational commitment between employees’ PCSR and innovation behavior. In the existing literature, some scholars have confirmed that PCSR can increase organizational commitment (Kong et al. 2019; George et al. 2021; Tran et al. 2021; Silva et al. 2023; Li and Qu 2023); Organizational commitment can positively influence innovation behavior (Nazir et al. 2018; Fauziawati 2021; Chen and Liu 2022; Chtioui et al. 2023). Li and Qu (2023) also studied combining internal CSR with organizational commitment and innovation behavior. However, these scholars often view organizational commitment as a singular construct or analyze it only in terms of affective commitment, lacking exploration of specific dimensions of organizational commitment and failing to integrate the multiple mediating relationships between employees’ PCSR, affective commitment, normative commitment, continuance commitment and employee innovation behavior.

This study finds that affective commitment, normative commitment and continuance commitment all positively impact employee innovation behavior and simultaneously play multiple mediating roles between PCSR and innovation behavior. The impact of employees’ PCSR on their innovation behavior can be understood within the S-O-R theoretical framework, where organizational commitment acts as an internal psychological process that explains individuals’ perception of their environment and forms specific behaviors (Wu and Chen 2018; Ahmed and Tahir 2019; Lu et al. 2020), mediating between perception and behavior and linking the PCSR to employee innovation behavior. This study also reveals that while all three types of commitment play a partial mediating role, the contributions of different types of commitment to the mediating effects vary slightly. The mediating effects of continuance commitment are the most significant. The findings of this study further elucidate the critical mediating mechanisms through which PCSR translates into innovation behavior through employee organizational commitment. This research enriches and refines our theoretical understanding of the origins and influences of employee organizational commitment and innovation behavior, thereby augmenting the existing S-O-R literature.

Fourthly, this study, conducted within the context of Chinese organizations, provides valuable contributions to the literature on the impact of employees’ PCSR on their psychology and work behavior in Eastern cultural settings. Additionally, AI enterprises are marked by innovation (Dong et al. 2020). The selection of Shenzhen—a model city for AI industry development in China (Shenzhen Artificial Intelligence Industry Association 2024)—as the focus of this study provides significant representativeness. Using employees from Shenzhen AI enterprises as subjects for examining innovation behavior offers robust support for a deeper understanding of the formation and enhancement of innovation capabilities among contemporary Chinese firms.

5.2. Managerial Implications

These study results have managerial implications and can offer business managers valuable recommendations. First, this study illustrates how employees’ attitudes and behaviors are influenced by their PCSR. The importance of studying PCSR and encouraging managers to consider employees’ perspectives on CSR activities is underscored by the fact that people’s views on CSR vary. Currently, many companies regularly publish CSR reports, using the disclosure of CSR information to make their management more transparent (Escamilla-Solano et al. 2024). Nevertheless, organizations may employ more methodologies to enhance their employees’ PCSR. For instance, employees’ PCSR can be enhanced by publishing information about CSR on websites, integrating CSR into training materials, routinely announcing CSR activities and encouraging employees to participate in these programs.

Secondly, employees’ PCSR still comes from objective CSR activities. Therefore, if business managers want to enhance employees’ organizational commitment and innovative behavior through PCSR, they still need to pay attention to CSR activities. Consequently, business managers should raise their level of CSR awareness, take on their fair share of duties to all parties involved and cultivate a stellar reputation and corporate image. Moreover, management, day-to-day operations and business development goals should all actively include CSR. Business managers need to understand the strategic significance of carrying out CSR. The company’s board of directors must set up a CSR group to standardize supervision and evaluate implementation when necessary. Business managers should end short-sighted behavior, position the company as the main body of CSR and oppose the paradigm of decision-making that prioritizes short-term financial gain above long-term CSR.

This study confirmed the positive relationship between PCSR, organizational commitment and employee innovation behavior. This conclusion can also encourage business managers to stimulate employees’ organizational commitment while fulfilling CSR to convey positive mediating effects. Understanding the role of different types of organizational commitments is essential for maximizing the promotion of innovation behavior. Considering that the mediating effect of continuance commitment is the most significant, it also shows that if employees can serve the company for a long time, it will be more effective in corporate innovation. To achieve this, business managers should pay special attention to employees’ psychological well-being, closely integrate social responsibility fulfillment with organizational commitment and amplifying the positive impact of PCSR on employee innovation behavior through strengthening organizational commitment. Specifically, in strengthening affective commitment, companies should enhance care and trust for employees to foster emotional bonds between employees and the organization. Long-term talent strategies, better work conditions and fair reward programs should be implemented to enhance the company’s reputation and foster employee identification with the company, thereby increasing employee continuance commitment. As for enhancing normative commitment, efforts should be made to enhance employees’ sense of belonging to the company, align their values with organizational values, cultivate a sense of responsibility and promote innovative behavior.

6. Conclusions

Through empirical testing, this study demonstrates that PCSR significantly improves employee innovation behavior and affective, normative and continuance commitments. PCSR has a stronger impact on continuance commitment. Affective, normative and continuance commitments positively impact employee innovation behavior; however, normative commitments have a stronger influence than affective and continuance commitments. Furthermore, affective, normative and continuance commitments mediate the relationship between employee innovation behavior and PCSR. The mediating effects of continuance commitment are the most significant, indicating that the impact of PCSR not only directly influences organizational commitment but also indirectly affects innovation behavior through the mediation of organizational commitment.

This study’s findings improve CSR research at the micro level and fill in existing gaps in the literature by extending research on the multiple mediation mechanisms between employees’ PCSR and innovation behavior. It also provides a broader perspective for examining the intricate relationships between employee attitudes and behaviors and PCSR. In addition, in practice, it provides a new perspective for business managers to find new motivations and channels for employee organizational commitment and innovation behavior, which helps to enhance corporate competitiveness and sustainable development.

Some limitations of this study include its primarily cross-sectional research approach and the inability to conduct a comprehensive sampling survey of the entire population due to financial and resource constraints. Future research may utilize objective indicators and longitudinal methods and consider conducting larger sample surveys in different countries, regions and among various types of enterprises.