1 Introduction

Over the past few decades, corporate social responsibility and ethical practices have become increasingly popular and influential in the social and economic landscape (Chouaibi & Chouaibi, 2021: 443). Adopting a CSR strategy has become a critical concern for businesses worldwide. This approach helps companies respond to various stakeholders' demands and expectations and enhance their competitive advantage and overall performance (Handayani et al., 2017: 152). Also, to conduct business in a way that considers environmental, social, and economic issues, the role of CSR is vital (Dai et al., 2022: 1). CSR is a strategy used by companies to achieve a competitive advantage by integrating academic and commercial objectives while also fulfilling their social responsibilities. Companies can differentiate themselves from their competitors by demonstrating their commitment to social responsibility while achieving their business goals (Huo et al., 2022: 4711). In the past, businesses primarily prioritized making profits, but now they must also consider the environment as a significant aspect of their operations (Hussain et al., 2022: 76,865). Companies focusing on CSR can enhance their ability to create new environmentally friendly practices. Additionally, enhancing green innovation is a crucial aspect of a company’s CSR efforts for its stakeholders (Javeed et al., 2022a: 2).

Green innovation empowers companies not just to address nationwide institutional pressures (Abdul & Jiang, 2024: 2) and stakeholder pressures, but also to enhance overall performance (Ali et al., 2023: 41). Also, green innovation plays a fundamental role in achieving sustainable performance for businesses (Kraus et al., 2020: 2). As the world increasingly prioritizes sustainable development goals, major global companies like Apple, Google, and Accenture have shifted their focus towards implementing sustainable business practices. Furthermore, a significant majority of executives from companies listed on the S&P 500 index (more than 84%) see innovation as a critical factor in ensuring the success of their business (Le, 2022: 2). Companies are being directed towards green innovation as it is a crucial element in achieving both environmental protection and economic growth, among other green initiatives (Li et al., 2022a: 1). Green innovation behavior involves taking into account environmental concerns in addition to financial benefits, unlike traditional innovation behavior. This may include developing environmentally friendly products or technologies, also known as green product innovation and green technology innovation (Li et al., 2022b: 1). Green innovation refers to the development of environmentally friendly products and services that can improve the utilization of resources and reduce environmental pollution. As a result, governments, non-governmental organizations, and consumers support green innovation activities because of their potential to solve environmental problems (Liu et al., 2022a: 3682).

Previous studies have suggested that CSR plays a significant role in encouraging the development of green innovation (Dai et al., 2022; Huo et al., 2022; Hussain et al., 2022; Le, 2022). This study’s primary goal is to explore the impact of CSR on green innovation, considering the neutral, negative, and positive findings on this relationship. Also, it examines how contextual factors may modify the effects of CSR on green innovation. This research has identified three distinct factors that exert the CSR-green innovation relationship within this scope. These factors, namely industry, type of data, and region, are of utmost significance when examining the relationship between CSR and green innovation.

This research uses a meta-analytical method to examine the correlation between CSR and green innovation by analyzing 29 empirical studies from peer-reviewed journals. The meta-analysis shows a strong and meaningful relationship between CSR and green innovation. Moreover, similar patterns were identified in the relationship between CSR and green innovation in manufacturing and other industries. Also, the results suggest that studies relying on primary data show a stronger correlation between CSR and green innovation than those relying on secondary data. The findings also highlight regional disparities in the CSR-green innovation relationship, with studies conducted in North America showing the most significant impact.

This current research makes several valuable contributions to the existing literature. First, this study represents the initial meta-analysis that explicitly examines the impacts of CSR on green innovation. This research demonstrates a noteworthy and favorable effect of CSR on green innovation. Second, the results indicate that the relationship between CSR and green innovation remains consistent across various industries. It reveals that CSR’s vivacious and significant effect on green innovation is robust for manufacturing and other industries. Third, it emphasizes the subjective bias in the primary data and shows that this bias has been confirmed in studies that reveal the CSR-green innovation relationship. Fourth, by demonstrating the applicability of the natural resource-based view (RBV) theory to both firm and individual levels in fostering green innovation through CSR practices, the research extends the theoretical framework’s scope, offering insights into how resources and capabilities contribute to innovation across different organizational levels. Finally, it highlights the importance of government policies for CSR practices and green innovation. It shows the possible effects of adopted strategies and policies on firms/individuals in the context of studies examining the CSR-green innovation relationship.

2 Literature review and hypotheses development

2.1 Theoretical background

The study is based on natural resource-based view theory. In recent times, RBV has gained prominence as a significant method for evaluating an organization’s competitive edge. RBV assists organizations in recognizing their internal strengths and weaknesses, enabling them to formulate strategies that leverage their strengths and mitigate their weaknesses (Malhotra et al., 2024: 1). Hart (1995) argued that the natural environment would present significant limitations and obstacles, making it a crucial catalyst for the emergence and advancement of resources and capabilities within organizations (Quintana-García et al., 2022: 1331). According to the RBV theory, an organization's resources and capabilities are crucial in achieving a competitive advantage. Furthermore, the natural RBV theory expands upon this idea and suggests that firms can achieve long-term competitive advantage by addressing environmental issues (Kraus et al., 2020: 2).

The natural RBV theory advocates for corporate sustainable practices. It highlights the importance of protecting natural resources harmed by global warming. As a result, this theory supports the implementation of sustainable policies by businesses to encourage green innovation (Javeed et al., 2022b: 4). By adopting CSR practices, a company can enhance the development of its intangible resources. As a result, the organization can create better organizational capabilities that lead to long-lasting competitive advantages (Marco-Lajara et al., 2022: 1474). Researchers can measure a company’s performance by using the natural RBV theory, which involves focusing on the company’s environmental, social, and economic aspects related to CSR (Kraus et al., 2020: 2). Moreover, CSR provides companies with the opportunity to incorporate external knowledge, which helps in expanding their knowledge base. This leads to creating fresh ideas and developing robust networks and relationships with stakeholders (Marco-Lajara et al., 2022: 1474).

2.2 Corporate social responsibility and green innovation

CSR encompasses a range of managerial approaches aimed at mitigating the adverse effects of corporate activities on society while concurrently maximizing their beneficial outcomes (Espaillat et al., 2022: 676). CSR refers to the idea that companies have a responsibility to consider and address the needs and interests of different groups in society beyond just their shareholders, and this responsibility goes beyond what is legally required or outlined in union agreements (Jones, 1980: 59–60; Oduro et al., 2022: 185). CSR has been a significant force in the business world for some time. Its significance is continuing to increase as it gains more support from business models and standards (Leonard & McAdam, 2003: 28). Although the origins of CSR can be traced back several centuries, it is sufficient for executives today to have a general understanding of its development over the past 50 years or so (Carroll, 2015: 87). CSR encourages managers to make decisions about topics that are outside their area of expertise. Additionally, it reinforces the belief held by managers and business school professors that it is the legitimate responsibility of business executives to address societal issues (Freeman & Liedtka, 1991: 94). Advocates of CSR present two distinct viewpoints: one argues that good corporate behavior can be beneficial for business, while the other suggests that pursuing social good may require sacrificing some profits (Benabou & Tirole, 2010: 9). CSR is considered strategic when it results in significant business-related advantages for the company, specifically by providing support for essential business operations, which in turn enhances the firm's ability to achieve its goals effectively (Burke & Logsdon, 1996: 496).

Green innovation can be defined as a method of creating new technologies and production processes that aim to minimize environmental risks such as pollution and the negative impacts of resource exploitation. It involves developing and implementing new ways to produce goods and services that reduce environmental impact (Arici & Uysal, 2022: 286; Naveed et al., 2023: 757). Green innovation refers to using technology to make improvements that save energy, prevent pollution, or facilitate waste recycling. This can include designing products in an environmentally friendly way and managing corporate environmental practices (Aguilera-Caracuel & Ortiz-de-Mandojana, 2013: 365). Contrary to general innovation, green innovation entails more significant technological novelty and complexity. It requires substantial expertise from specialized individuals and relies on establishing enduring relationships through long-term relational contracts with customers and suppliers (Tang et al., 2023: 1044).

Nowadays, green innovation has become a critical asset for businesses as it helps them expand their market share and maintain their existence in the long term. A well-executed green innovation strategy enhances the market position of a business, draws in more customers, offers environmentally-friendly services, and provides a competitive edge over rivals (Karimi Takalo et al., 2021: 2). Effective green innovation enables companies to establish core strengths and elevate entry barriers by providing unique products, marketing preventative technologies, and seizing emerging business prospects (Qadeer et al., 2024: 61). Also, businesses can use green innovation to improve their environmental performance and meet the global standards for environmental conservation (Chang & Chen, 2013: 1058). Encouraging sustainable economic growth and improving societal welfare heavily relies on embracing green innovation (Sun et al., 2023: 1). Companies that are the first to introduce new environmentally-friendly products can benefit from being a pioneer in the market, allowing them to charge higher prices for their green products (Chen, 2008: 535). Companies might prioritize green innovation initiatives over other projects involving innovation (Chen et al., 2023: 2). To attain enduring and sustainable progress, companies also must prioritize the implementation of green innovation (Xue and Wang, 2024: 1834). Nevertheless, certain companies lack the motivation for green innovation due to the burden some limitations of expensive investments, significant risks, and positive and negative impacts on external factors (Wu et al., 2023: 644).

A CSR strategy can encourage companies to allocate more resources toward developing eco-friendly technologies and products that meet stakeholders’ expectations. The success of green innovation is influenced by how much importance a company places on fulfilling its social responsibilities. Therefore, the CSR strategy devised by top executives can potentially drive the company towards green innovation (Liu et al., 2022b: 4). According to Porter and Kramer (2006), CSR serves as a vital tool for promoting green innovation, gaining a competitive advantage, and enhancing shareholder value. Martinez-Conesa et al. (2017) claimed that CSR can increase the likelihood of investments in research and development, which in turn can drive innovation in environmentally friendly or “green” technologies. Hull and Rothenberg (2008) argued that when companies engage in CSR initiatives, it can result in the development of environmentally sustainable innovation. Furthermore, Hong et al. (2020) claimed that engaging in CSR initiatives can catalyze the promotion of green innovation.

According to the natural RBV theory, organizations have the ability to gain a competitive edge by implementing diverse environmental tactics, including initiatives focused on environmental preservation, sustainable product innovation, and resource conservation. Green innovation functions as a crucial asset for organizations and a constantly developing capability. Through the utilization of this framework, firms have the potential to surpass rivals by enhancing the influence of CSR on green innovation and their ecological achievements (Bonsu et al., 2024: 224). As per the natural RBV theory, organizations are advised to address environmental obstacles through the cultivation of distinct resources and competencies. Consequently, utilizing CSR and green innovation competencies as resources may aid in alleviating the adverse effects of industrial operations, subsequently improving economic performance (Simmou et al., 2023: 2). Moreover, in accordance with the natural RBV theory, green investment and CSR are considered as significant assets and competencies within organizations, which can support the implementation of green innovation (Le & Ferasso, 2022: 1034). Several studies also support the correlation between CSR and green innovation (Dai et al., 2022; Ma et al., 2023; Zhou et al., 2023). Based on the literature mentioned earlier, the first hypothesis is as follows:

H1

CSR is positively related to green innovation. Increases in CSR lead to higher levels of green innovation.

2.3 Moderating effects of industry, data type, and region

Certain factors unique to each research can strengthen the relationship between CSR and green innovation. Three specific conditions are especially significant in this connection: industry, data type, and region. The use of technology can significantly improve the environmental performance of various industries. One example is the positive impact that manufacturing operations based on process technologies can have on the environment, as these processes often involve recycling and other environmentally friendly practices (Yu et al., 2022: 713). As people become increasingly concerned about the impact of environmental challenges, the manufacturing industry has shifted its focus towards developing more environmentally-friendly technology. It is crucial for companies, especially those in manufacturing industries that generate high levels of pollution, to take responsibility for protecting the environment (Xu et al., 2022: 2). Manufacturing companies are more likely to cause pollution, so they tend to follow environmental regulations more closely and place a greater emphasis on developing green innovation technologies (Javeed et al., 2022b: 5). Manufacturing companies have been commended by environmentalists for adopting CSR practices and incorporating eco-friendly thinking into their operations, in order to achieve both economic and environmental sustainability (Javeed et al., 2022a: 6).

A manufacturing company's financial and environmental performance can be significantly influenced by two crucial factors—CSR and the advancement of green innovation (Xu et al., 2022: 1). Moreover, manufacturing companies are facing a higher level of scrutiny compared to other industries, making it more essential for them to provide precise information regarding corporate social issues (Javeed et al., 2022b: 9). The theoretical framework of the natural RBV contributes to elucidating the rationale behind the investment of manufacturing companies in sustainability, emphasizing environmental considerations as a pivotal internal competency. This viewpoint substantiates the organization's capacity to attain a sustainable competitive edge. Furthermore, the natural RBV theory posits that enterprises have the potential to leverage their interactions with the natural surroundings to establish persistent competitive benefits (Hsu et al., 2023: 459). The natural RBV theory also recognizes the necessity for organizations to engage with various stakeholders, including suppliers, customers, regulators, and environmental groups, in order to effectively oversee natural resources and uphold sustainable procedures (Johan et al., 2023: 398). Companies involved in manufacturing exhibit a more conscientious approach towards the environment by diminishing or abstaining from the exhaustion and deterioration of natural resources, thus safeguarding their integrity for an extended duration (Dzage et al., 2024: 3). Therefore, manufacturing companies can obtain more excellent knowledge to enhance their green innovation through the implementation of CSR. In summary, the second hypothesis is as follows:

H2

CSR’s effect on green innovation is stronger in manufacturing firms than in other industries.

Another specific condition whose moderator effect was examined is primary/secondary data. Primary data is gathered and analyzed by the researcher or research team. In primary data analysis, the researcher plans and designs the study, collects data from participants, and analyzes the collected data (Mongan, 2013: 374). On the contrary, secondary data may have been previously published, or maybe the original data was collected by someone else (Church, 2002: 32). Secondary data analysis involves reanalyzing existing data or information that was originally collected by someone else or for a different purpose than the current one being considered (Mongan, 2013: 374). Surveys can be influenced by self-reporting bias, where individuals may intentionally over- or under-report information due to social desirability reasons or memory errors such as forgetting or rounding dates (Partin et al., 2003: 867).

On the other hand, according to the principle of transparency, the data disclosed by companies to the public should not contain misleading information, which indicates that the secondary/published data is more objective. However, participants may express the opinion that there are more initiatives regarding CSR and/or green innovation than what is currently known. Therefore, primary data obtained through surveys or interviews are more likely to be subjective. Supporting this, Zehrer et al. (2017) found that subjective primary data diverges from the findings derived from objective secondary data. Similarly, McCracken et al. (2001) examined hospital managers’ subjective perceptions (primary data) alongside the hospital's objective financial performance data (secondary data) and found that the correlations between these subjective and objective measures varied.

Resources in the context of a natural RBV perspective can be classified as either physical or non-physical in essence. To create value, these resources need to possess a level of scarcity, and additionally, they demonstrate a quality of being “tied semi-permanently” to the firm (Guillaume & Julia, 2021: 5). Managers and employees are considered vital internal assets for organizations from this standpoint. As a result, their perspectives and attitudes towards CSR and green innovation carry considerable significance in the domain of primary data. To summarize, individuals tend to exaggerate CSR and/or green innovation activities due to subjectivity in primary data, while secondary data is more transparent and likely to present more accurate CSR and/or green innovation activity information. As a result, it is assumed that the relationship between CSR and green innovation is higher in research that uses primary data. Thus, the third hypothesis is as follows:

H3

CSR’s effect on green innovation is stronger in studies using primary data than those using secondary data.

Azam et al. (2022) stated that the green theory advocates achieving regional, national, and international environmental sustainability. Green innovation stems from the green theory and establishes a connection between a company’s operations and the environment. According to Su and Zhong (2022), the extent of local economic and social development, the surrounding environmental conditions, and government intervention significantly impact enterprises operating in various regions. As a result, there are noticeable variations in enterprises’ geographical and political characteristics regarding their approach to communicating social responsibility. For example, Du et al. (2019) concluded that Chinese industrial enterprises exhibit notable regional imbalances and disparities in their effectiveness regarding green innovation. In particular, the eastern and western regions demonstrate higher efficiency than the central regions, while the northeast region exhibits the lowest efficiency level.

Yu et al. (2022) claimed that regional environmental regulations could influence the impact of green innovation on CSR fulfillment. Regional environmental regulation encompasses the various restrictive laws, policies, measures, and implementation procedures local governments implement to promote environmental protection. Steurer et al. (2012) asserted that Western European governments, specifically those of Anglo-Saxon and Scandinavian nations, exhibit a notably higher level of commitment to promoting corporate social responsibility (CSR) compared to governments in Central and Eastern European (CEE) countries. However, natural resources constitute some of the most crucial economic assets for any nation. RBV may disregard these resources, but this subject encompasses the natural RBV (Rehman et al., 2024: 15,304). Numerous developing economies currently encounter severe environmental challenges with substantial repercussions for the environment and human existence. Through the natural RBV perspective, CSR and green innovation capabilities can be employed as resources to alleviate the adverse effects of industrial operations (Simmou et al., 2023: 2). To summarize, the last hypothesis is that:

H4

The impact of CSR on green innovation differs from one continent to another.

The depiction of the theoretical framework described earlier, along with the interconnections and impacts among the variables, is illustrated in Fig. 1 as the research model.

Fig. 1
figure 1

Research model

Full size image

Figure 1 underscores the direct correlation between CSR and green innovation, indicating that an escalation in CSR practices leads to a corresponding increase in green innovation. Also, it illustrates how three variables—industry, data type, and region—moderate the relationship between CSR and green innovation. Lastly, it unveils that within this study, the relationship between CSR and green innovation aligns with the natural RBV theory.

3 Methodology

3.1 Sampling

This study uses the PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) approach to determine the sample. The PRISMA process involves four steps: identification, screening, eligibility, and inclusion (Moher et al., 2010). These steps are further explained in Fig. 2.

Fig. 2
figure 2

The PRISMA approach

Full size image

Initially, certain aspects, such as relevant keywords, appropriate databases, and the search period, are determined in the identification process. The study has recognized and distinguished two groups of keywords: (“corporate social responsibility,” “corporate social responsiveness”) and (“green innovation”, “green product innovation”, “green process innovation”, “green technology innovation”, “environmental innovation”, “eco-innovation”, “ecological innovation”). Sustainability encompasses more than just environmental issues; it also includes social sustainability. However, this study will focus solely on the environmental aspect of green innovation concept and will not examine social sustainability. Therefore, it excludes the concept of “sustainable innovation”. A search is conducted on Google Scholar, Web of Science, and Scopus databases, using keywords to find published papers. To clarify, a search for indexed documents in the databases will be conducted by the end of March 2023. One thousand six hundred sixty-four papers from Google Scholar, 548 from Web of Science, and 841 from Scopus were obtained. Once the process of combining the three data sets to eliminate duplicate entries was completed, the resulting collection of papers comprised 1586 distinct samples.

Second, the screening process involves applying a quality criterion to ensure the quality of this research. Specifically, only peer-reviewed journal articles are included in the research sample. Thus, books, book chapters, proceedings, and thesis are excluded. As a result, a sample of 947 articles is obtained.

Third, the articles are subjected to a thorough examination based on six specific eligibility criteria: (1) The primary emphasis of the study should be directed towards green and environmental concerns while excluding studies that solely examine social sustainability. (2) CSR practices and green innovation are firm-level variables. Therefore, while individual opinions are considered in the primary data analysis, they specifically pertain to the CSR and green innovation practices of the organization. Thus, only firm-level studies are included. (3) The research needs to be quantitative and include well-defined constructs and measurements while addressing CSR and green innovation concurrently. Thus, papers with unclear construct definitions, mathematical modeling papers, and case studies are not included. (4) The research must include the calculation of Pearson's correlation between CSR and green innovation. Papers that solely provide β coefficients will not be considered, as meta-analytic outcomes derived from Pearson’s correlations are deemed to be more precise. (5) Only the study that provides comprehensive details and is published in a journal with high impact is considered when a dataset is utilized in multiple studies. (6) This study did not include articles whose full text could not be accessed. Access to the Pearson correlation coefficient is crucial to this meta-analysis research. Nine hundred eighteen studies failed to fulfill all four criteria and were excluded.

At the end of the PRISMA process, a final valid sample of 29 studies was obtained. In the research sample, the earliest published research dates back to 2020, and 29 articles have been published between 2020 and 2023, demonstrating the growing importance and popularity of the CSR-green innovation relationship as a contemporary and trending topic.

3.2 Measures

The study employed three categories of variables: independent variables, dependent variables, and moderator variables.

Independent variable: The two keywords representing the independent variable are “corporate social responsibility” and “corporate social responsiveness.”

Dependent variable: Green Innovation encompasses various innovative approaches to tackle environmental issues. These approaches include innovations such as “green innovation”, “green product innovation”, “green process innovation”, “green technology innovation”, “environmental innovation”, “eco-innovation”, and “ecological innovation”.

Moderators: This meta-analysis includes three moderators. Industry, data type, and region are three specific conditions considered particularly important to connect CSR and green innovation. This study coded the industry variable as manufacturing sector 1 and other sectors 0. It also coded the other moderator variable, primary data, as 1 and secondary data as 0. Finally, it coded the region variable as follows: (1) Asia, (2) Europe, (3) Africa, (4) North America, and (0) Mixed.

3.3 Publication bias

Assessing publication bias involves examining both visual and statistical aspects. Initially, the fail-safe N is used to evaluate the potential presence of publication bias. This analysis assumes that unpublished studies have an effect size of zero. It aims to determine the minimum number of publications with a zero effect size required to obtain a non-significant p-value (Rosenthal, 1979). If only a few unpublished studies can challenge the meta-analysis, the results may not be reliable. This study’s fail-safe N for the relationship between CSR-green innovation is 9601, suggesting no significant publication bias. Furthermore, the rank correlation test conducted by Begg and Mazumdar yielded an insignificant outcome (p > 0.1). Finally, a visual representation of a funnel plot is shown in Fig. 3. The symmetrical funnel plot proves that publication bias does not significantly affect research findings.

Fig. 3
figure 3

The funnel plot

Full size image

3.4 Meta-analytic procedures

The average effect size is employed when a study contains multiple effect sizes. To account for measurement error, the initial correlation coefficient in each study is adjusted by multiplying it with an attenuation factor. This factor is determined by taking the square root of the two constructs’ CR (composite reliability) (Hunter & Schmidt, 2004). Once the correlations have been corrected, they undergo an additional transformation into Fisher’s z values to reduce the impact of variance of sampling.

The Q and I2 are employed to assess the variation in effect size across studies, indicating heterogeneity. A notable Q value suggests dissimilarity among the sampled studies. The I2 value signifies the degree of heterogeneity, with higher values indicating more significant variation. Increased heterogeneity levels suggest moderators’ potential existence (Li et al., 2022c: 6).

4 Results

4.1 Effect size

The random effects model examines how CSR affects green innovation by accounting for variations in the actual effect size across different studies. This is necessary because each study has unique characteristics influencing the effect size. Also, Garrido-Ruso and Aibar-Guzmán (2022) and Katz et al. (2022) used the random effects model to analyze the zero-order effects of two variables. The requirement for the random effects model is confirmed by considerable heterogeneity, as evidenced by the high Q and I2 values. Thus, it is deemed suitable for the analysis (Borenstein et al., 2010).

This study involved the estimation of 29 effect sizes. Figure 4 presents a forest plot that summarizes these effect sizes. The author’s name, Fisher’s Z values, publication year, and the relative weight of each article are depicted in the forest plot. Most studies demonstrate distributions centered around the mean effect size value of 0.446. As per Cohen’s (1988) effect size classification, the correlation coefficient and sample size calculations indicate a medium and positive association between CSR and green innovation. Thus, H1 is supported.

Fig. 4
figure 4

The forest plot

Full size image

The Q statistic, with a value of 1648.623 and 28 degrees of freedom (df), indicates significance at a 95% confidence interval (p = 0.000). Furthermore, the heterogeneity of the data can be assessed by examining the I2 value. In this study, the I2 value was 98.302. Further investigation is warranted due to the considerable heterogeneity observed in the CSR-green innovation link among the sampled studies, as evidenced by the high Q and I2 values.

To assess the reliability of research findings through sensitivity analyses, a one-sample-removed analysis is conducted on the final sample of 29 studies. This technique involves systematically eliminating 1 research at a time from the sample to examine the potential impact of each study on the results (Kepes et al., 2013). This process is repeated 29 times. In this way, this study was able to assess whether a single study had a significant impact on the overall relationship between CSR and green innovation. According to the results of this analysis, the effect sizes for the CSR-green innovation relationship range from 0.416 to 0.462, providing further evidence of the robustness of the research findings.

When the dataset is heterogeneous, the ANOVA approach can be employed (Lipsey & Wilson, 2001). In this study, three categorical variables, namely industry, data type, and region, were examined. Consequently, ANOVA (subgroup) analyses were utilized and outlined in the following sections.

4.2 Effect of industry

The researchers investigated whether the data was obtained from the manufacturing industry. Studies focused on such datasets were categorized as 1, while studies unrelated to the manufacturing industry were classified as 0. Subsequently, the ANOVA approach was employed to analyze the impact of these categorical variables. Two distinct models, fixed and mixed models, allow for subgroup analyses. Borenstein et al. (2009) emphasized the importance of examining the significance of the QTOTAL statistic in the fixed-effects model. If the test statistic yields a significant result, it is crucial to consider the outcomes of the mixed-effects model. In line with the fixed-effects model results, all Q values were found to be significant (QMANUFACTURING = 263.751; QOTHER = 1151.176; QW = 1414.928; QB = 233.695; QTOTAL = 1648.623). As QTOTAL demonstrated significance, the mixed-effects model was employed to investigate potential effects, and the corresponding findings are presented in Table 1.

Table 1 ANOVA results by industry
Full size table

Table 1 presents the mean effect size values of 0.463 and 0.431 for the manufacturing and other categories, respectively. In both groups, the effect was statistically significant (p = 0.000). Consequently, it can be concluded that CSR positively influences the green innovation levels of both groups. The QB value, which indicates whether the effect size varies between the groups, was determined to be 0.117 (with 1 degree of freedom, p = 0.733). This outcome suggests no significant difference in the mean effect size values between the subgroups. Thus, H2 is not supported.

4.3 Effect of data type

A subgroup analysis was performed to investigate if the impact differs across data types. The data type variable consists of two categories, namely primary and secondary. If the study utilizes primary data, it is indicated as “1”; otherwise, it is symbolized as “0.” Based on the results obtained from the fixed effects model, all Q values were statistically significant. The significant Q values were QPRIMARY = 404.874, QSECONDARY = 253.304, QW = 658.178, QB = 990.445, and QTOTAL = 1648.623. Since the QTOTAL value was substantial, the potential effects were examined using the mixed-effects model, and the outcomes are presented in Table 2.

Table 2 ANOVA results by data type
Full size table

Table 2 demonstrates that the mean effect size values for primary and secondary data groups were 0.616 and 0.169, respectively. The effect was statistically significant across all groups (p = 0.000). Consequently, it can be concluded that CSR effectively influences green innovation levels for all data types. However, the QB value was 36.622 (with 1 degree of freedom, p = 0.000). Therefore, there is a significant difference in mean effect size values based on the data type. In simpler terms, CSR impacted green innovation differently for each data type. It is concluded that the CSR-green innovation relationship is higher in studies using primary data than in studies using secondary data. Thus, H3 is supported. The main reason may be individuals’ subjective evaluation of CSR-green innovation activities during the primary data collection.

4.4 Effect of region

A subgroup analysis was performed to examine if the effect size differs across different regions. The variable “region” comprises five categories: Asia, Europe, Africa, North America, and mixed. This variable indicates the geographical region where the research took place. To illustrate, when surveying participants in China, the region was classified as Asia. Conversely, if the study was conducted in Germany, the region was classified as Europe. Studies conducted in regions such as Italy and Morocco were classified as mixed. Based on the outcomes of the fixed-effects model, significant Q values were observed for all categories except North America (p = 1.000) and Africa (p = 1.000). The Q values for each category were as follows: QASIA = 1466.249, QEUROPE = 41.952, QMIXED = 33.474, QW = 1541.676, QB = 106.947, and QTOTAL = 1648.623. Since the QTOTAL value was significant, potential effects were further analyzed using the mixed-effects model, and the corresponding results can be found in Table 3.

Table 3 ANOVA results by region
Full size table

Table 3 presents the mean effect size values for various regions. Specifically, the effect sizes were 0.434, 0.474, 0.568, 0.685, and 0.394 for Asia, Europe, Africa, North America, and Mixed. The effect was significant in all groups (p = 0.000) except for Europe (p = 0.095). Hence, it can be concluded that CSR significantly influences green innovation across all regions except for Europe. The QB value was 12.433 (with 4 degrees of freedom, p = 0.014). Consequently, there is a significant difference in the mean effect size values among the different regions. In other words, CSR impacts green innovation differently across all areas. Thus, H4 is supported. This impact is highest in North America. This region is followed by Africa, Asia, and the group with mixed samples regarding effect size. In North America, the research was conducted in the Dominican Republic, which has a thriving economy with responsible green innovation strategies. The sample comprises 158 agricultural firms with many resources and management assets, high turnover levels, intensive information technologies and market intelligence use, and good HRM practices. Therefore, this may be the main reason for the high effect size.

5 Discussion

By analyzing data gathered from 29 empirical studies, this research investigates the influence of CSR on green innovation and concludes a positive correlation between CSR and green innovation (H1). This finding is consistent with the findings of previous studies (Dai et al., 2022; Fang, 2020; Kraus et al., 2020; Novitasari & Tarigan, 2022; Xu et al., 2022; Yu et al., 2022). Moreover, the present study examines the influences of industry, data type, and region as moderating factors. The findings from the research model are illustrated in Fig. 5. This visual representation not only depicts the relationship between CSR and green innovation but also highlights the involvement of three moderator variables in shaping this relationship.

Fig. 5
figure 5

Results of research model

Full size image

The findings show that the CSR-green innovation relationship does not differ by sector. The CSR-green innovation relationship was similar for manufacturing and other industries (H2). The fact that the CSR-green innovation relationship is identical for manufacturing and other sectors can be attributed to the increasing awareness of CSR and green innovation practices in almost all sectors today and the critical role of firms in increasing their environmental and social contributions through these practices. For example, Huo et al. (2022) claimed that given the current circumstances of stricter environmental regulations, all industries in China have no alternative but to prioritize enhancing the effectiveness of green innovation to attain economic growth and ecological preservation simultaneously. Similarly, Breitbarth and Rieth (2012) emphasized that although CSR is difficult to integrate and manage, many companies in all sectors are adopting CSR for competitive advantage, moving towards incorporating these practices into their corporate structure/culture and business strategies/activities, and even CSR activities are important in professional football organizations.

The result of the study on the second moderator variable indicates that CSR had a different impact on green innovation for each data type. Besides, the results reveal that the CSR-green innovation relationship is higher in studies using primary data than in studies using secondary data (H3). This result may be because primary data is reflected differently than it is, and secondary data is more realistic because it is shared with the public. Here, there is a concern about subjective bias in the primary data collection process. The subjective bias does not mean that all primary data is subject to bias. For example, a woman you survey at the airport may misrepresent her age. This is indicated by subjective bias, which is inevitable but not at a level that can affect the entire study. Indeed, Stimson and Marans (2011) classify secondary data, such as official statistics and reports, as objective, while primary data, such as surveys and interviews, are classified as more subjective. In the current research, studies that investigated the CSR-green innovation relationship using primary data proved that this relationship is higher.

As another moderating variable, this study investigated whether the CSR-green innovation relationship differs by region. According to the results, the impact of CSR on green innovation varies among different regions, and the most significant effect is seen in studies conducted in North America (H4). This result may be because environmental policies, strategies, and restrictions adopted by countries directly affect firms’ activities, and firms adapt their innovative technologies to their processes and operations by considering ecological concerns. These policies and strategies set by countries will likely have joint effects on firms and consumers in the context of CSR and green innovation. According to Steurer (2010), CSR initially emerged as a neo-liberal idea to reduce government regulations. However, Steurer observed that it has evolved into a more progressive strategy focused on societal collaborative regulation. Therefore, regional differences will likely emerge due to different policies adopted, especially regarding the CSR-green innovation relationship and other environmental factors. Moreover, there are differences in CSR practice even among countries in the same region. For example, Steurer et al. (2012) argued that governments in Western European nations, specifically those of Anglo-Saxon and Scandinavian origins, are considerably more proactive in promoting Corporate Social Responsibility than governments in Central and Eastern European countries. Also, the high effect size of the study conducted in the North American region may be because the sample includes large-scale firms that use information technologies effectively and have advanced innovation processes. Primary data were obtained by interviewing these firms.

6 Conclusion and future directions

This study utilizes meta-analysis to explore the relationship between CSR and green innovation. This research presents evidence supporting a solid and favorable relationship between CSR and green innovation by combining data from 29 different empirical studies. The study also examined the CSR-green innovation relationship in the context of three moderator variables: industry, data type, and region. The relationship between CSR and green innovation exhibited similarities across the manufacturing sector and other industries. Also, the findings indicate that the CSR-green innovation relationship is more robust in studies utilizing primary data than those using secondary data. The findings reveal that the CSR-green innovation relationship differs across various regions, with the most significant impact observed in studies carried out in North America.

This study has certain limitations. First, a causal relationship between CSR and green innovation cannot be confirmed because a correlation coefficient was used to measure effect size. Second, green process, product, and technology innovation are coded as green innovation proxies. In future research, the relationship between these three variables and CSR can be investigated separately. Third, only articles in three databases were analyzed in the study. The number of databases and types of publications can be increased in future research. Fourth, studies emphasizing only the environmental dimension of sustainability were examined, ignoring its social and economic dimensions. In future research, sustainability can be discussed in all its dimensions. Fifth, there's a restriction in the sample regarding the number of studies available for various regions. Specifically, there's only one study each included in the sample for North America and Africa. Finally, three moderator variables were used in the study. Researchers may use moderator variables such as firm size, sample size, and country in future studies.

6.1 Theoretical implications

This study makes significant contributions to the existing literature on CSR-green innovation. First, this study establishes noteworthy advancements in CSR and green innovation research by pioneering a meta-analytical inquiry to systematically evaluate the correlation between CSR and green innovation. While there has been an increase in the number of studies examining the relationship between CSR and green innovation in the past decade, there is a scarcity of review articles considering CSR as the underlying cause of green innovation. Moreover, no studies have employed a quantitative literature analysis to explore the correlation between CSR and green innovation. The lack of information motivates author to conduct a statistical analysis of the current inconclusive results and provide a more conclusive determination. The robust findings demonstrate that CSR plays a significant role in shaping green innovation. Therefore, this research enhances the relationship between these two areas: CSR and green innovation.

Second, according to Javeed et al. (2022b), the natural RBV theory advocates for businesses to adopt sustainable policies that foster green innovation. However, Dai et al. (2022) claimed that companies seek to establish and uphold their credibility by harmonizing their corporate values, endeavors, and approaches with those of the community, aiming to earn acceptance and preserve it over time. The findings of this study reveal that CSR practices at both the firm and individual levels promote green innovation. In this respect, this result extends the natural RBV theory to apply to both levels. Moreover, the existence and validity of the CSR-green innovation relationship for manufacturing firms and firms in other industries extend this theory.

6.2 Managerial implications

This study has important policy and management implications. First, this study asserts that CSR may be a tactical decision businesses make to advance green innovation. Also, CSR initiatives have been proven to increase green innovation for manufacturing and other companies significantly. Therefore, businesses of any industry may embrace CSR practices. However, this research concluded that primary data studies had a more extensive CSR-green innovation relationship than secondary data studies. This proves that the information obtained from firm employees and managers reveals a more significant CSR-green innovation relationship. The opposite may be true in reports on the firm’s CSR practices. At this point, it is crucial that all employees adopt the CSR practices of the firm and that employees are aware of these practices.

Secondly, policymakers should understand that when companies choose to use CSR practices, it will enhance their ability to be environmentally friendly, ultimately improving their overall green innovation. Companies will benefit in a good way by doing CSR activities. Furthermore, this effect will be more substantial in countries/regions prioritizing being environmentally friendly. So, it is suggested that government officials use either money or other methods to make the country more environmentally friendly. For example, the people who make decisions can make rules to make the state support green causes better, make the country more appealing for investing in renewable energy, and create good ways to measure how well green things are doing.