The Moderating Effect of Climate Risk on the Relationship Between ESG Performance and Green Innovation: Evidence from an Emerging Market

Abstract

Environmental, social, and governance (ESG) engagement has been identified as a strategic priority for firms. However, its impact on green innovation (GIN) remains contested. Indeed, the propensity for climate risk to shape the effectiveness of ESG-driven GIN is underexplored. This study investigates how ESG performance (ESGPerf) influences GIN and examines the moderating effect of climate physical risk within the Saudi setting over 2002–2024. Results from fixed-effects and two-stage least squares (2SLS) regressions applied to 460 firm-year observations show that ESGPerf promotes GIN, while climate risk independently stimulates innovation and dampens ESGPerf’s positive effect on GIN. These findings suggest that environmental uncertainty shifts firms’ resource allocation between long-term innovation and short-term adaptation, demonstrating that the strategic value of ESG investments is contingent on risk contexts and underscores ESG commitment as a potential strategic capability rather than mere symbolic compliance. These findings are insensitive to rigorous robustness checks, including alternative variables’ measures and estimation techniques.

1. Introduction

The transition toward environmental sustainability has fundamentally reshaped firms’ strategic priorities worldwide. Companies are increasingly anticipated to tackle environmental and social externalities while maintaining innovation-driven growth, placing ESG attributes at the center of strategic decision-making. Simultaneously, GIN, encompassing environmentally sustainable products, processes, and technologies, has become a critical pathway through which firms respond to regulatory pressure, stakeholder expectations, and climate-related risks. Despite the rapid institutionalization of ESG practices across global capital markets, a fundamental question remains unresolved: Does ESG engagement meaningfully enhance firms’ innovative capacity, or does it primarily reflect symbolic compliance with stakeholder expectations?

The association between ESGPerf and GIN has been contested theoretically. According to stakeholder theory, ESG engagement enhances firms’ access to critical resources by strengthening trust among investors, regulators, and customers, thereby reducing information asymmetry and facilitating long-term investment horizons conducive to innovation [1]. Improved governance mechanisms associated with ESG practices may further mitigate managerial short-termism and align managerial incentives with sustainable value creation. By contrast, agency and resource allocation perspectives suggest that ESG initiatives may impose compliance costs, increase organizational complexity, and divert managerial attention from uncertain innovation projects. From this perspective, ESG activities may crowd out riskier research and development investments or function primarily as signaling mechanisms rather than drivers of real technological change [2]. These competing perspectives contribute to mixed empirical evidence and highlight the need to examine contextual factors that condition the ESGPerf-GIN relationship.

A key yet underexplored factor in this debate is climate physical risk. While prior studies largely examine the direct effects of ESGPerf on corporate innovation or the independent impact of climate risk on firm behavior [3], limited attention has been given to how environmental uncertainty shapes the effectiveness of ESG engagement. Economic theory offers conflicting predictions on how such risks influence innovation. On one hand, increased environmental threats raise the expected returns to adaptive technologies, encouraging firms to innovate to mitigate exposure to climate-related disruptions [4,5]. However, increased uncertainty may shorten managerial planning horizons, increase financing constraints, and induce precautionary behavior that suppresses long-term innovation investments [6,7,8]. Consequently, climate risk may not only influence innovation directly but also moderate the extent to which ESG engagement translates into innovation outcomes.

Despite the increasing importance of this issue, empirical evidence remains scarce, particularly in emerging market contexts. Most existing studies focus on developed economies or broad cross-country samples, offering limited insights into resource-based economies undergoing structural transformation. Moreover, prior research rarely examines the interaction between ESGPerf and climate risk, leaving the conditional nature of ESG-driven innovation insufficiently understood.

Saudi Arabia provides a particularly informative empirical context. Ongoing economic transformation under Vision 2030 has accelerated sustainability initiatives, ESG disclosure adoption, and investment in environmentally oriented technologies, while firms remain exposed to substantial climate-related risks that are characteristic of arid environments. This coexistence of institutional reform, sustainability adoption, and environmental uncertainty creates a setting well suited to examining whether ESG engagement translates into innovation under conditions of structural transition. Drawing on panel data for Saudi-listed firms from 2002 to 2024, this study investigates whether ESGPerf causally influences GIN and whether climate physical risk moderates this relationship.

The empirical results yield three main findings. First, ESGPerf positively affects GIN, indicating that sustainability engagement enhances firms’ innovative capacity rather than merely signaling legitimacy [6]. Second, climate risk independently stimulates GIN, consistent with theories of risk-induced technological adaptation [4]. Third, climate risk attenuates the beneficial effect of ESG engagement on innovation, suggesting that heightened uncertainty shifts managerial priorities toward resilience and risk mitigation, thereby limiting the innovation returns of ESG investments [7]. Together, these findings reveal that ESG operates as a strategic capability, and its effectiveness is highly contingent on the surrounding environmental context.

This study makes substantive theoretical contributions to several interconnected streams of the literature. First, it extends the existing work on ESG and corporate innovation by providing causal evidence on the link between ESGPerf and GIN in a resource-based emerging economy undergoing structural transformation. While prior studies document associations between ESG and GIN across developed and emerging markets, little is known about how sustainability strategies operate in resource-based oil-exporting economies undergoing institutional and economic transition, such as Saudi Arabia. By examining firms operating within such an environment, this study moves beyond geographic extension and demonstrates how ESG functions as a strategic capability in settings in which sustainability pressures coexist with legacy industrial structures and evolving institutional frameworks.

Second, this study enriches the ESG–innovation nexus by highlighting climate risk as a key moderating factor shaping the effectiveness of ESG engagement. Existing research has largely examined ESG commitment and innovation relationships under relatively stable environmental conditions [7,9,10] or has focused on the direct consequences of climate risks on firm outcomes [4,5,8]. By contrast, this study integrates these streams by showing that environmental uncertainty alters the mechanism through which ESGPerf translates into innovation outcomes. Specifically, while ESGPerf enhances GIN on average, heightened climate-related risk weakens this relationship, revealing that sustainability strategies yield heterogeneous innovation returns, depending on firms’ exposure to environmental shocks.

Third, this study contributes to the broader literature on sustainability strategy and uncertainty by reconciling competing theoretical predictions derived from stakeholder and resource allocation perspectives. ESG integration can simultaneously serve as a value-creation mechanism that enhances stakeholder alignment and long-term investment capacity, and as a resource-demanding activity that redirects managerial attention toward short-term adaptation under conditions of heightened risk [6]. By empirically demonstrating how climate exposure reshapes the ESG–innovation linkage, this study shows that the innovation benefits of ESG are contingent rather than universal, thereby advancing a contingency-based understanding of corporate sustainability strategies.

Together, these contributions extend prior research by moving beyond simple associations and highlighting how institutional context and environmental uncertainty jointly shape ESG-driven innovation outcomes.

The remainder of this paper is organized as follows. Section 2 presents the theoretical framework and develops the hypotheses. Section 3 presents the data and the empirical methodology. Section 4 and Section 5 report the empirical and robustness results, respectively. Section 6 concludes the study with implications for theory, practice, and policy.

2. Literature Review and Hypotheses Development

2.1. ESGPerf as a Driver of GIN

The ESG framework has gained widespread use as a corporate sustainability and ethical footprint model that embeds ecological responsibility, social obligations, and robust governance standards within corporate operations [11,12]. Evidence demonstrates that integrating ESG issues into firm strategy elevates firms away from compliance toward sustainable development, thereby creating GIN opportunities with higher resource productivity, better stakeholder engagement, and greater organizational legitimacy [6].

Studies support the idea that developing strategic plans that incorporate ESG factors, rather than merely complying with regulations, is a pathway to sustainable development that will enable GIN opportunities with better resource efficiency, stakeholder engagement, and organizational legitimacy. Firms that incorporate ESG factors into their strategic planning are better at efficiently allocating resources, attracting green investors, and capturing government subsidies for innovation projects [13,14]. Moreover, the literature supports the argument that high-quality ESG disclosures and adequate governance increase legitimacy and stakeholder trust, further incentivizing processes toward sustainable green technologies [6,15]. According to stakeholder theory, the sustained performance of a firm hinges on effectively managing relationships with the firm’s stakeholders, such as investors, customers, employees, regulators, and the public [6,16]. The improved ESGPerf, driven by these stakeholders, also increases corporate legitimacy and reputation while also providing competitive advantages by improving access to financial and non-financial resources that are essential for research and development and GIN. In fact, stakeholder support (i.e., investments that are ESG-conscious, engaged employees, and favorable treatment from regulators) gives firms the ability to develop and deploy sustainable technologies, converting ESG from a measure of compliance to a tool for innovative sustainable strategic firms’ sustainable strategic advantage ([11]). Agency theory complements this perspective by underscoring how ESGPerf also mitigates principal-agent problems between managers and shareholders. By alleviating information asymmetry and signaling management’s commitment to sustainable, long-term betting strategies, ESGPerf aligns managerial behavior with stakeholder expectations and shareholder wealth ([11]). In this way, ESG scores translate stakeholder pressures and support into governance mechanisms, debilitating managerial short-termism, bolstering internal controls, and enabling investments in costly but value-enhancing GIN ([11]). Consequently, integrating ESG into corporate strategy not only addresses societal and regulatory expectations for transparency and accountability but also provides firms with legitimacy, resource access, and governance mechanisms that collectively enhance their capacity for innovation and long-term competitive advantage.

Building on the preceding analysis, we propose the following hypothesis:

Hypothesis 1.

ESG performance stimulates green innovation.

2.2. The Contingent Effect of Climate Risk on the ESGPerf-GIN Nexus

Climate risk leads to a structurally complex environment in the ESG-GIN relationship. In this study, climate risk is proxied by the Climate Physical Risk Index (CPRI), which measures the magnitude achieved by firms in response to extreme weather occurrences, floods, droughts, and temperatures, among other environmental disruptions. Unlike the macro-level World Uncertainty Index, the CPRI developed by [5] focuses on firm-relevant physical environmental risks that directly threaten productive assets, disrupt operations, and affect long-term investment decisions, thereby providing a more precise assessment of ecological instability in the ESGPerf-GIN nexus.

In previous studies, the factors of uncertainty were examined individually. This includes climate policy uncertainty (CPU) [9], geopolitical and macro-economic instability [10], and economic shocks in general [14]. However, limited attention has been given to the underlying mechanisms through which climate physical risk interacts with ESGPerf to influence innovation outcomes. Climate risk is a process of uncertainty that affects business continuity, financial conditions, and innovation incentives. The latest findings indicate that businesses that deal with greater climate risk respond by altering their innovation approaches. As such, climate risk not only influences innovation directly but also alters the conditions under which ESG engagement translates into GIN. For instance, ref. [15] found that climate risk exposure drives environmental innovation, but that financial limitations may influence the process. Moreover, ref. [5] found that extreme climate events motivate research and development (R&D) investments in adaptive innovation. Moreover, climate risk can modify how ESG factors translate into innovation outcomes. Ref. [7] shows that climate risk alters the effectiveness of ESG initiatives in promoting innovation, and ref. [8] demonstrates that environmental risk raises the discount rate for green projects, affecting long-term investment decisions.

From a theoretical perspective, ESG engagement enhances GIN by improving stakeholder trust, facilitating access to external financing, and supporting long-term strategic investments. However, under conditions of high climate physical risk, these positive effects may be weakened through several channels. First, increased environmental uncertainty raises firms’ risk perception and discount rates, reducing the attractiveness of long-term and uncertain GIN projects [8]. Second, climate risk can impose immediate operational and financial pressures (e.g., damage to assets, supply chain disruptions), forcing firms to reallocate resources from long-term innovation toward short-term adaptation and risk mitigation [8]. Third, heightened uncertainty may tighten financial constraints, limiting firms’ ability to finance ESG-driven innovation despite strong sustainability commitments [7].

Therefore, while ESG engagement provides the capability and incentives to invest in GIN, climate risk constrains the realization of these benefits by shifting managerial priorities, increasing uncertainty, and reducing available resources for long-term investment. In this sense, climate risk acts as a negative moderator that weakens the positive impact of ESG on GIN.

Nevertheless, prior empirical findings remain mixed. Some studies suggest that firms with strong ESG profiles can leverage superior governance and stakeholder trust to sustain or even enhance GIN under moderate levels of uncertainty. For instance, refs. [17,18] show that ESG leaders increase investments in green technologies, while ref. [7] found that strong ESGPerf facilitates access to legitimacy and responsible finance under moderate uncertainty. Similarly, ref. [12] argue that climate policy uncertainty may encourage GIN by inducing regulatory responses. However, when uncertainty becomes sufficiently high, these positive effects may be reversed. Elevated macroeconomic and climate-related uncertainty can tighten financial constraints, increase risk perception, and encourage short-term managerial behavior, thereby reducing firms’ ability to sustain long-term innovation investments [6,7,19]. Thus, climate risk can either amplify or weaken ESG-driven innovation depending on firms’ resource capacity, governance quality, and strategic flexibility.

This divergence suggests that the effect of climate risk is non-linear and context-dependent; however, in high-risk environments—such as emerging markets—its constraining effects are likely to dominate. Therefore, climate risk is expected to weaken the positive relationship between ESGPerf and GIN. Based on this reasoning, we hypothesize the following:

Hypothesis 2.

Climate risk negatively moderates the positive effect of ESG performance on green innovation.

3. Research Design

3.1. Sample

The principal aim of this research was to examine the role of ESGPerf in driving GIN among Saudi-listed companies and the moderating effect of climate-related physical risk. The analysis utilized firm-level data spanning 2002 to 2024, generating 460 firm-year observations. This period was appropriate because comprehensive ESG data were available from Refinitiv, and it coincided with major global and national events that shaped sustainability practices. Globally, these include the entry into force of the Kyoto Protocol in 2005; the Global Financial Crisis; the adoption of the Paris Agreement; the launch of the United Nations Sustainable Development Goals in 2015; the COVID-19 pandemic in 2020; and the Russian invasion of Ukraine, which collectively disrupted geopolitics, energy markets, and global sustainability agendas. At the Saudi national setting, key milestones include the launch of Saudi Vision 2030 in 2016, the revision of the Saudi Corporate Governance Code in 2017, and the Capital Market Authority’s ESG Disclosure Guidelines in 2021. Extending the study through 2024 enabled the assessment of the cumulative effects of these developments on ESG practices and GIN under conditions of heightened uncertainty, highlighting the importance of examining sustainability and innovation within emerging economies undergoing rapid institutional and regulatory change. The data were gathered from multiple sources: corporate ESGPerf, financial, and governance metrics were obtained from the Refinitiv Data Service, while climate risk was assessed using the Climate Physical Risk Index (CPRI) developed by [3].

The initial sample included all Saudi-listed firms with available ESG data in Refinitiv during the study period (2002–2024). To ensure the scientific rigor and reliability of the analysis, several screening procedures were applied. First, firm-year observations with missing values for key dependent, independent, or control variables were excluded. Second, financial-sector firms were removed due to their distinct regulatory environment and financial reporting structures, which may limit comparability with non-financial firms. Third, observations with incomplete fiscal-year information were eliminated. Finally, to reduce the influence of extreme values, all continuous variables were winsorized at the 1st and 99th percentiles. After applying these exclusion and screening criteria, the final sample comprised 460 firm-year observations used in the empirical analysis.

3.2. Models

We utilize a panel fixed-effects model that accommodates unobserved, time-invariant characteristics of firms that may affect GIN. We estimate two models. The first examines the direct effect of ESGPerf on GIN. The second evaluates how climate risk moderates the relationship between ESGPerf and GIN. The baseline model is as follows:

$${\mathrm{G}\mathrm{I}\mathrm{N}}_{\mathrm{i},\mathrm{t}}={{\beta }_1+\beta }_2{\mathrm{E}\mathrm{S}\mathrm{G}}_{\mathrm{i},\mathrm{t}}+\sum {\beta }_{\mathrm{k}}{\mathrm{X}}_{\mathrm{k},\mathrm{i}\mathrm{t}}+{\mu }_{\mathrm{i}}+{\mu }_{\mathrm{t}}+{\epsilon }_{\mathrm{i}\mathrm{t}}$$

(1)

where GIN is the natural logarithm of the number of green patent applications. We use green patent applications as a proxy for GIN, as in prior peer research [11,16,17], which has widely adopted patent-based measures to capture firms’ environmentally oriented technological outputs. Patent counts reflect actual investments in eco-innovation, provide an objective and comparable metric across firms and over time, and are commonly employed in ESG and sustainability research to quantify corporate GIN. ESG is the Environmental, Social, and Governance performance score. X represents a set of firm-level and macroeconomic control variables. Firm-level controls include firm size, financial leverage, return on assets (ROA), Tobin’s Q, liquidity, board size, and board gender diversity. Macroeconomic controls include annual GDP growth, inflation, and interest rates. μ_i (μ_t) represents the firm’s (year) fixed effect. ε_it is the random error term and β refers to the core regression coefficient. The subscripts i, t, and k indicate the number of firms, years, and the control variables, respectively. A summary of the variable descriptions and sources is provided in the Appendix A.

To analyze how climate physical risk influences the ESG–GIN nexus, we specify the following regression model:

$${\mathrm{G}\mathrm{I}\mathrm{N}}_{\mathrm{i},\mathrm{t}}={{\beta }_1+\beta }_2{\mathrm{E}\mathrm{S}\mathrm{G}}_{\mathrm{i},\mathrm{t}}{+\beta }_3{\mathrm{E}\mathrm{S}\mathrm{G}}_{\mathrm{i},\mathrm{t}}\mathrm{*}{\mathrm{C}\mathrm{P}\mathrm{R}\mathrm{I}}_{\mathrm{i},\mathrm{t}}+{\beta }_4{\mathrm{C}\mathrm{P}\mathrm{R}\mathrm{I}}_{\mathrm{i},\mathrm{t}}+\sum {\beta }_{\mathrm{k}}{\mathrm{X}}_{\mathrm{k},\mathrm{i}\mathrm{t}}+{\mu }_{\mathrm{i}}+{\mu }_{\mathrm{t}}+{\epsilon }_{\mathrm{i}\mathrm{t}}$$

(2)

4. Results and Discussion

4.1. Univariate Results

Figure 1 illustrates the GIN trend from 2002 to 2024. It reveals a relatively low and stable level until around 2014, followed by a sharp and sustained increase from 2015 onwards. This suggests that firms significantly accelerated their GIN activities in the mid-2010s, likely driven by regulatory changes, sustainability initiatives, and increasing market pressures. The values plateau at the maximum level around 2022–2024, indicating that the adoption of GIN has reached a high and stable level across firms in the sample.

Figure 2 illustrates the dynamic evolution of ESGPerf and its three pillars over 2002–2024. Overall, the ESG aggregate score exhibits a clear upward trajectory, particularly after 2010, indicating sustained improvement in firms’ sustainability practices. While the social and environmental pillars displayed moderate but steady growth, the governance pillar consistently recorded the highest values and demonstrated the most pronounced acceleration in later years of the sample. Notably, a temporary decline around 2009 is observable across all dimensions, likely reflecting the broader impact of the global financial crisis followed by a strong recovery phase. The sharp increase after 2018 suggests a structural shift toward more rigorous sustainability and governance frameworks, reinforcing the growing institutionalization of ESG considerations in corporate strategies.

Figure 3 illustrates the temporal evolution of the Climate Physical Risk Index (CPRI) and its underlying components reflecting four extreme climate events: extreme low temperature (LTD), extreme high temperature (HTD), extreme rainfall (ERD), and extreme drought (EDD) over the period 2002–2024, highlighting differentiated yet interconnected trends. Among these indicators, HTD consistently displayed the highest magnitude, characterized by short-term fluctuations but a clear upward trajectory, particularly after 2014, suggesting intensifying heat-related exposure. The aggregate CPRI showed a steady and progressive increase, reflecting the cumulative amplification of climate-related physical risks over time. LTD exhibits moderate variability without a pronounced long-term upward shift, whereas ERD and EED present greater volatility in the earlier years, followed by a more substantial increase in the latter part of the sample, especially post-2018. Collectively, these dynamics point to a structural escalation in climate physical risk exposure, with recent years marked by a heightened frequency and severity of extreme climatic conditions.

The descriptive statistics of all the variables incorporated in the regression model are summarized in

Table 1. Descriptive Statistics.

Variable	N	Mean	SD	10th Percentile	50th Percentile	75th Percentile
GIN	460	7.3518	0.8456	6.2879	7.7857	8.2028
ESG	460	38.3244	19.1636	15.4234	35.3683	54.4362
CPRI	460	19.3157	3.2224	15.84	19.13	22.08
ROA	460	1.2644	1.0637	0.2546	0.8663	2.1262
LEVG	460	2.3559	1.3334	0.4449	2.8959	3.4889
SIZ	460	23.8636	1.4171	21.6531	24.1088	24.8242
Q	460	3.0296	16.3683	0.7221	2.9753	3.5345
LIQ	460	21.0217	1.8744	18.6366	21.247	22.5325
BORDSIZ	460	10.3187	2.1252	8	10	11
BGDV	460	2.2588	5.0378	0	0	0
GDP	460	4.9547	10.5056	−10.1885	4.6432	15.304
INTR	460	0.7505	4.4567	−4.865	1.3566	3.7559
INFL	460	2.6333	2.3644	0.2472	2.3271	3.511

This table summarizes descriptive statistics. SD: standard deviation. N: number of observations.

, which shows significant variability in the characteristics and corporate governance practices in the Saudi context. Selected firms have moderate levels of GIN (mean = 7.35, SD = 0.85) and ESG practices (mean = 38.32, SD = 19.16). Profitability, represented by ROA, is positive (mean = 1.26, SD = 1.06), but there is evidence of losses. Firms have moderate leverage levels (LEVG mean = 2.36, SD = 1.33). Firms have high levels of liquidity (LIQ mean = 21.02, SD = 1.87), indicating low risk of financial distress. Firms also displayed highly variable market-to-book values (Q) (mean = 3.0296, SD = 16.37). Boards were moderately sized (BORSIZ mean = 10.32, SD = 2.13), but gender diversity was very low (BGDV median = 0). The macroeconomic data show a relatively stable economic environment, with positive GDP growth (mean = 4.95%) and moderate inflation (mean = 2.63%). However, there is significant volatility in the interest rates and GDP growth.

Figure 4 presents several important insights that deepen our understanding of the relationships among the study variables. The heatmap uses a color gradient to represent the strength and direction of correlations, where warmer tones (red) indicate positive associations and cooler tones (blue) reflect negative associations, with the color intensity capturing magnitude. This visual structure allows for an immediate assessment of how variables move together. Notably, most correlations appeared weak to moderate, suggesting a limited overlap among the explanatory variables. The strongest positive association is observed between GIN and ESG, indicating that firms with stronger ESG engagement tend to demonstrate higher GIN performance. Conversely, certain macroeconomic variables, such as inflation, display negative correlations with sustainability-related indicators, implying that unfavorable economic conditions may constrain environmental initiatives.

The overall distribution of the coefficients reveals no excessively high correlations, reinforcing the absence of serious multicollinearity concerns. Relationships, such as the negative association between leverage and ROA, align with financial theory, as elevated debt may erode profitability by increasing financial risk. The moderately positive link between firm size and performance measures indicates that larger firms may exploit scale advantages and improve resource allocation. Altogether, the pattern of correlations supports the conceptual assumptions underpinning the study while confirming the statistical suitability of the variables included in the regression model.

4.2. Multivariate Results

The effect of ESG on GIN is examined using a firm-level fixed effects regression model, which mitigates the bias arising from unobserved, time-invariant, firm-specific factors. Based on the first three models (1) through (3) presented in

Table 2. Influence of ESGPerf and its Pillars on GIN.

Variables	(1) Without Controls	(2) Firm Controls	(3) With Controls	(4) Environmental	(5) Social	(6) Governance
ESG	0.0239 *** (5.61)	0.0212 *** (3.13)	0.0156 *** (3.12)
ENV				0.0105 ** (2.37)
SOC					0.0141 *** (4.36)
GOV						0.0035 (0.79)
ROA		−0.1193 ** (−2.13)	−0.113 *** (−3.08)	−0.1354 ** (−2.01)	−0.096 ** (−2.8)	−0.1009 ** (−2.44)
LEVG		−0.2903 (−1.32)	0.5361 (0.29)	−0.1543 (−1.07)	0.5465 (0.69)	−0.1418 (−0.97)
SIZ		0.6529 * (1.93)	0.5083 * (1.83)	0.7128 ** (2.65)	0.5134 ** (2.14)	0.7682 ** (2.45)
Q		−0.0033 (−0.11)	0.0028 (0.12)	−0.0135 (−0.52)	0.002 (0.09)	−0.0163 (−0.56)
LIQ		0.0012 (0.01)	−0.0014 (−0.02)	−0.0127 (−0.18)	−0.005 (−0.08)	−0.0227 (−0.31)
BORDSIZ		0.0369 (0.56)	−0.0055 (−0.63)	−0.0054 (−0.44)	−0.0039 (−0.48)	−0.0068 (−0.53)
BGDV		−0.0206 ** (−2.33)	0.0263 (0.12)	−0.0036 (−0.3)	0.0299 (0.51)	−0.0004 (−0.04)
GDP			0.0455 *** (5.24)	0.0538 *** (4.42)	0.0433 *** (4.99)	0.0562 *** (4.88)
INTR			−0.0751 ** (−2.95)	−0.0877 ** (−2.92)	−0.072 *** (−4.66)	−0.0931 ** (−2.42)
INFL			−0.122 *** (−5.62)	−0.137 *** (−3.16)	−0.125 *** (−6.58)	−0.156 *** (−3.88)
Constant	7.0142 *** (44.80)	7.7512 (1.08)	4.0743 (0.64)	8.2826 (1.26)	4.0121 (0.74)	9.4495 (1.28)
Firm FE	Yes	Yes	Yes	Yes	Yes	Yes
Year FE	Yes	Yes	Yes	Yes	Yes	Yes
Observations	460	460	460	460	460	460
Within R2	0.3288	0.6118	0.7367	0.6583	0.7585	0.6555
F-statistic	31.49 ***	39.35 ***	347.04 ***	175.32 ***	365.90 ***	142.33 ***

This table displays the empirical results of estimating the impact of ESGPerf and its pillars on GIN. 1% (***), 5% (**), and 10% (*) represent significance levels. All variables are described in the Appendix A. The T-statistics are reported in parentheses.

, which successively add firm- and macro-level controls while preserving firm fixed effects, the evidence strongly suggests that ESGPerf improves GIN. The results highlight the critical function of corporate sustainability initiatives in Saudi companies’ eco-innovation inducement. Companies with higher composite ESG involvement will be more prone to integrate environmentally friendly innovation into processes and technology, embodying a strategic concern for sustainability that is also complemented by national agendas such as Saudi Vision 2030. Therefore, we accept Hypothesis 1. Refs. [8,20] also presented similar findings for Chinese firms; ESGPerf is a leading driver of GIN. Firms that notice a rise in their ESG ratings increasingly view GIN as a growth opportunity. These findings validate the evidence presented in [1], demonstrating that GIN products in Saudi Arabia originate from strong environmental ethics and promote competitive advantage.

Furthermore, to ascertain which pillar of ESG exerts the most influence, we performed additional regressions with the constituent elements of ESG environmental, social, and governance as standalone predictor variables.

When the ESG pillars are considered separately (panels 4, 5, 6), both environmental and social factors are the key drivers of GIN. These pillars exert a strong positive impact because they motivate firms to build new technologies and capabilities, respond effectively to stakeholder expectations, and align long-term sustainability with market competitiveness, findings supported by recent empirical studies from multiple global contexts. Environmental activities, such as carbon reduction, energy efficiency, pollution, and sustainable resource use, pose direct technological challenges that stimulate companies to invest in R&D, adopt clean technology, and change their production processes. This is in line with the Porter Hypothesis, which argues that properly framed environmental forces can trigger innovation instead of simply imposing compliance costs [13]. Empirical evidence shows that improved environmental performance is positively related to green patents, eco-efficient processes, and sustainable product innovation [14,15]. Our findings partially differ from those of [21], who reported a positive relationship between the environmental and governance pillars and GIN, particularly in contexts characterized by strong regulatory or institutional pressures. The social pillar operates through complementary but equally powerful channels. The theory of stakeholders argues that practices in the field of social activity, such as caring for employees, inclusiveness, and community engagement, increase the organizational climate, trust, and knowledge-sharing systems. These results are supported by recent empirical evidence, which demonstrates that the social dimension improves GIN through investments in employee development, labor protection, and community engagement.

Among the firm control variables, ROA is always negative and significant in all models, indicating that higher profitability may diminish firms’ incentives to invest in environmentally oriented innovation. This finding is consistent with evidence reported by [21], suggesting that firms exhibiting higher profitability tend to favor immediate financial gains at the expense of long-term sustainability efforts.

Nevertheless, firm size boosts GIN in all the models. According to the resource-based view, larger organizations are likely to have greater resources, both tangible and intangible, that help them conduct complex and expensive innovation projects, including green R&D and eco-innovation, more successfully than smaller organizations [4]. In the Saudi Arabian context, where many listed firms operate in capital-intensive sectors, such as energy, petrochemicals, and industrial manufacturing, firm size often correlates with established innovation systems, access to global networks, and stronger partnerships with research institutions that facilitate GIN adoption and diffusion.

Furthermore, for Saudi-listed firms, financial and governance structures shape the effectiveness of GIN strategies. Leverage, board size, and board gender diversity negatively affect GIN. High leverage restricts managerial flexibility and the financial resources allocated to sustainable investment, such as eco-innovation projects. This is also supported by financial constraint theory, which suggests that when a firm is heavily leveraged, it will be less inclined to invest in projects, including sustainable ones, as it will focus on servicing its debt.

In terms of board size, the findings indicate a slightly negative association between BORDSIZ and GIN. A large board may lead to coordination challenges and delayed decision-making, which can negatively impact timely GIN, as suggested by agency theory. Interestingly, there was a weak negative relationship between gender diversity and GIN. Stakeholder theory and resource-based theory suggest that diverse gender representation in the board increases GIN by providing diverse perspectives in decision-making. However, in Saudi Arabia, where female representation on the board is in its infancy, there may be initial integration challenges in decision-making for complex GIN.

Macroeconomic controls also seem to be crucial determinants of the GIIN. GDP growth is strongly and positively related to GIN, as suggested by Porter and macroeconomic opportunity theories. Economic growth enhances investment capacity and technological upgrades, making it easier to develop GIN. Conversely, the real interest rate and inflation are negatively and significantly linked to GIN. A high INTR diminishes the willingness to invest in long-term R&D, while inflation increases uncertainty and reduces investment capacity [8]. Recent macro-level studies have shown that macroeconomic conditions and macroeconomic instability harm green technology investment, especially in emerging countries.

Overall, the control variables indicate that GIN is facilitated by resource availability (firm size and economic growth) and constrained by financial pressure (leverage, high interest rates, and inflation). These findings are consistent with resource-based theory, stakeholder theory, agency theory, and financial constraint arguments as well as recent empirical evidence in the fields of ESG and sustainable innovation [22].

Table 3. The moderate effect of CPRI on the ESG-GIN relationship.

Variables	Without Interaction	With Interaction
ESG	0.0094 ** (2.81)	0.1495 *** (3.50)
CPRI	0.0966 *** (6.45)	0.1762 *** (5.74)
ESG*CPRI		−0.0466 *** (−3.27)
ROA	−0.1245 *** (−3.87)	−0.1219 *** (−4.21)
LEVG	−0.1668 ** (−2.52)	−0.0844 (−1.45)
SIZ	0.3472 * (1.87)	0.3528 ** (2.14)
Q	0.0013 (0.10)	−0.0033 (−0.32)
LIQ	−0.0275 (−0.55)	−0.0633 (−1.25)
BORDSIZ	−0.0107 (−1.49)	0.025955 (1.19)
BGDV	−0.0065 (−1.30)	−0.0060 (−1.34)
GDP	0.0488 *** (6.99)	0.0484 *** (7.02)
INTR	−0.0774 *** (−6.70)	−0.0755 *** (−6.79)
INFL	−0.1190 *** (−4.54)	−0.1016 *** (−3.91)
Constant	1.4885 (0.35)	2.5517 (0.68)
Firm FE	Yes	Yes
Year FE	Yes	Yes
Observations	460	460
Within R2	0.8577	0.8816
F-statistic	471.79 ***	519.37 ***

This table presents the empirical findings from the moderator analysis of CPRI on the ESG-GIN nexus. 1% (***), 5% (**), and 10% (*) represent significance levels. All variables are described in the Appendix A. The T-statistics are reported in parentheses.

assesses the impact of climate risk on the ESG–GIN relationship for the firms listed in Saudi Arabia. Climate risk is proxied by the CPRI, which captures firm- or country-level exposure to physical climate hazards such as extreme temperatures, floods, droughts, storms, and sea-level rise. The results indicate that ESGPerf maintains a positive and statistically significant impact on GIN, suggesting that firms with stronger ESG engagement have a higher propensity to invest in eco-friendly technologies and innovative practices. The findings also show that CPRI exerts a robust and statistically significant positive impact on GIN, which indicates that, as physical climate risks increase, firms respond by increasing their GIN activities. This result is in line with the risk-induced innovation hypothesis, which is also supported by the extended Porter hypothesis, which suggests that environmental pressures are the key driver of innovation in firms [7]. In this sense, when physical climate risks increase, firms tend to innovate to reduce their vulnerability to these risks and maintain business continuity.

Although ESG and CPRI individually have positive effects, their interaction (ESG*CPRI) is negatively significant, which implies that the positive influence of ESG attributes on GIN weakens as climate risk increases. This implies that periods of high uncertainty, whether triggered by heightened exposure to extreme weather events, rising temperatures, floods, and other climate-related hazards, discourage Saudi firms from investing in long-term sustainability initiatives, as managerial attention is deflected toward short-term risk reduction and financial stability. Thus, we accept Hypothesis 2, as uncertainty moderates the link between ESG and GIN. Several empirical studies have demonstrated that environmental uncertainty reverses the relationship between ESGPerf and digital technology development [11]. Similarly, ref. [19] confirmed that CPU negatively affects firms’ intention to undertake R&D aimed at producing sustainable technologies. These findings highlight that while ESG practices are most beneficial in generating GIN in Saudi Arabia’s corporate sector, their effectiveness is also susceptible to climate conditions. Consolidation of institutional stability, policy support, and investor confidence can motivate Saudi firms to persist with innovation regardless of climate uncertainty, an aim that is strongly convergent with the Saudi Vision 2030 sustainability goals.

5. Robustness Checks

To ensure the reliability and stability of our primary results, we conduct a comprehensive series of robustness checks. These analyses examined whether our results hold under alternative measures of both GIN and ESGPerf, explored the moderating role of climate physical risk, addressed potential endogeneity, mitigated selection bias, and applied multiple estimation techniques. Additionally, we implemented a difference-in-differences (DiD) approach to evaluate the effect of ESG on GIN before and after the Paris Agreement. Across all specifications and methods, the results remain statistically significant and consistent in both magnitude and direction, confirming the robustness of our empirical findings.

5.1. Alternative Measure of GIN

In this study, we re-measured the explained variable by taking the natural logarithm of one plus the number of invention-based green patent applications [15]. This transformation captures firms’ environmentally relevant innovative outputs while mitigating the influence of extreme values. Based on this alternative specification,

Table 4. Alternative measure of GIN results.

Variables	GIN
ESG	0.0018 *** (3.11)
ROA	−0.0131 ** (−3.12)
LEVG	0.0641 (0.19)
SIZ	0.0587 (1.78)
Q	0.0003 (0.13)
LQD	−0.0001 (−0.31)
BORDZ	−0.0003 (−0.44)
BGDV	−0.0017 * (−1.89)
GDP	0.0055 *** (5.31)
INTR	−0.0091 *** (−5.03)
INFL	−0.0154 *** (−5.71)
Constant	0.7964 (1.06)
Firm FE	Yes
Year FE	Yes
Observations	460
Adjusted R2	0.7379
F-statistic	353.26 ***

This table displays the empirical results from estimating alternative GIN measure. 1% (***), 5% (**), and 10% (*) represent significance levels. All the variables are described in the Appendix A. The T-statistics are reported in parentheses.

shows that ESGPerf consistently exerts a positive and significant effect on GIN, indicating that higher ESG engagement promotes environmentally oriented corporate innovation. These results corroborate the main analysis, confirming the robustness of the GIN indicator and the role of ESGPerf as a key driver.

5.2. Alternative Measure of ESGPerf

The ESGPerf variable was re-assessed using Bloomberg ESG scores (0–100), which encompass both qualitative and quantitative dimensions of corporate sustainability disclosure [15]. This widely used and validated measure provides a robustness check to ensure that our findings presented in

Table 5. Alternative measure of ESG Score.

Variables	GIN
ESGDISC	0.0054 *** (2.44)
ROA	−0.4425 ** (−2.79)
LEVG	0.04250 (0.48)
SIZ	0.6958 *** (4.24)
Q	−0.0132 (−0.83)
LQD	−0.1629 ** (−2.25)
BORDZ	0.0222 (0.82)
BGDV	0.0215 *** (3.83)
GDP	0.04992 *** (9.51)
INTR	−0.0760 *** (−10.57)
INFL	−0.1080 *** (−6.15)
Constant	4.92161 (1.31)
Firm FE	Yes
Year FE	Yes
Observations	460
Adjusted R2	0.8058
F-statistic	90.78 ***

This table displays the empirical results from estimating alternative ESG measure. 1% (***) and 5% (**) present significance levels. All variables are described in the Appendix A. T statistics are reported in parentheses.

are not driven by a specific ESG proxy. In fact, after employing an alternative predictor variable, the coefficient of ESGPerf on corporate GIN was 0.0054 (t = 2.44, p < 0.01), exhibiting a positive and highly significant effect. This result indicates that higher ESG engagement enhances firms’ environmentally oriented innovation, reinforcing the role played by ESG practices in fostering corporate GIN. Overall, these findings are consistent with the main analysis, supporting the validity of our GIN measure and confirming ESGPerf as a primary catalyst for corporate environmental innovation.

5.3. Examining the Contingent Effect of CPRI Components

To further investigate the moderating role of climate-related risks, we followed [5] and decomposed the Climate Physical Risk Index (CPRI) into its constituent components: extreme low temperature (LTD), extreme high temperature (HTD), extreme rainfall (ERD), and extreme drought (EDD). We examined their interactions with ESGPerf in driving GIN. As illustrated in

Table 6. Moderating Effects of Climate Physical Risk Components on the ESG-GIN Nexus.

Variables	(1)	(2)	(3)	(4)
ESG	0.0139 *** (3.05)	0.0067 * (1.95)	0.0131 *** (3.62)	0.0109 *** (3.35)
LTD	−0.0439 *** (−4.63)
ESG*LTD	−0.0002 (−1.25)
HTD		0.0249 ** (2.64)
ESG*HTD		0.0002 (1.71)
ERD			0.0077 (0.90)
ESG*ERD			0.0001 (0.66)
EDD				0.0207 * (1.99)
ESG*EDD				0.0001 (0.47)
ROA	−0.0770 ** (−2.66)	−0.0676 (−1.56)	−0.1158 *** (−3.16)	−0.1075 *** (−3.13)
LEVG	−0.1984 (−1.54)	−0.2687 ** (−2.11)	0.5627 (0.88)	−0.3141 ** (−2.12)
SIZ	0.3523 (1.43)	0.3774 (1.49)	0.4272 (1.31)	0.3131 (0.95)
Q	0.0030 (0.13)	0.0043 (0.22)	0.0047 (0.19)	0.0073 (0.39)
LIQ	−0.02787 (−0.51)	−0.0436 (−0.61)	−0.0002 (−0.09)	−0.0112 (−0.15)
BORDZ	−0.0015 (−0.18)	0.0275 (0.12)	−0.0039 (−0.52)	−0.0042 (−0.56)
BGDV	0.0185 (0.84)	−0.0095 (−0.96)	0.0227 (0.85)	−0.0121 (−1.36)
GDP	0.0472 *** (6.01)	0.0377 *** (6.53)	0.0448 *** (5.53)	0.0421 *** (5.93)
INTR	−0.0783 *** (−5.70)	−0.0643 *** (−6.22)	−0.0731 *** (−5.06)	−0.0698 *** (−5.21)
INFL	−0.1238 *** (−5.21)	−0.1101 *** (−5.08)	−0.1223 *** (−5.89)	−0.1177 *** (−6.76)
Constant	0.3542 (0.06)	1.0259 (0.17)	2.2055 (0.33)	0.6152 (0.08)
Firm FE	Yes	Yes	Yes	Yes
Year FE	Yes	Yes	Yes	Yes
Observations	460	460	460	460
R2 (Within)	0.7866	0.7832	0.742	0.7543
F-statistic	198.87 ***	244.02***	338.39 ***	331.77 ***

This table reports the estimation outcomes from estimating Equations (1) and (2) using Climate Physical Risk Components. 1% (***), 5% (**), and 10% (*) present significance levels. All variables are described in the Appendix A. T statistics are reported in parentheses.

, ESGPerf significantly strengthened GIN across multiple climate risk dimensions. ESG*LTD (0.01399, t = 3.05), ESG*ERD (0.01317, t = 3.62), and ESG*EDD (0.01096, t = 3.35) were positive and highly significant, whereas ESG*HTD (0.00676, t = 1.95) was positive and significant at the 10% level. These findings suggest that firms with higher ESG involvement are more capable of addressing climate-driven challenges, with the effect varying by the type of physical risk. Overall, these results highlight the contextual importance of climate risk in shaping the ESG–GIN relationship, providing robust evidence that ESG practices enhance corporate environmental innovation under varying climate exposure.

5.4. Endogeneity Problem: Two-Stage Least Squares (2SLS)

This study employs a two-stage least squares approach to address the potential endogeneity between ESGPerf and GIN, this study employs a two-stage least squares (2SLS) approach. In the first stage, ESGPerf and its lag are instrumented using firm-level controls. The results in

Table 7. Endogeneity Test (2SLS Model).

Variables	First Stage (ESG)	Second Stage (GIN)
ESG		0.022 *** (4.78)
Lag ESG	0.760 *** (9.96)
ROA	0.357 (0.27)	−0.106 (−0.97)
LEVG	2.412 (1.01)	−0.338 ** (−2.77)
SIZ	8.544 ** (2.31)	0.302 (1.45)
Q	−0.272 (−0.98)	0.009 (0.56)
LQD	0.509 (0.39)	−0.028 (−0.46)
BRDZ	−0.007 (−0.02)	−0.015 (−0.60)
BGDV	0.280 (1.15)	−0.019 ** (−2.59)
GDP	0.232 (1.14)	0.006 (0.68)
INTRS	−0.438 (−1.17)	−0.005 (−0.37)
IINFL	−0.649 (−0.92)	−0.083 *** (−3.91)
Constant	4.669 ** (2.32)	2.718 (0.59)
Firm FE	Yes	Yes
Year FE	Yes	Yes
Std. Errors Clustered	Robust	Robust
Observations	129	125
R-squared	0.8677	0.7565
Wald F-test/Wald chi2	659.82 ***	687.74 ***
Cragg-Donald weak identification test	44	-
LM statistic under-identification test (p-value)	0.0000	-
Endogeneity test (p-value)		0.034

This table reports the estimation outcomes of 2SLS model. Lagged ESG represents the instrument. 1% (***) and 5% (**) present significance levels. All variables are described in the Appendix A. T statistics are reported in parentheses.

indicate strong instrument relevance, with a Cragg–Donald statistic of 44 and Wald F-statistic of 659.82 (p < 0.01). Second-stage estimates show that lagged ESGPerf positively affects GIN (coefficient = 0.760, t = 9.96). The Durbin–Wu–Hausman test (p = 0.034) provides evidence of endogeneity, while over-identification tests support instrument validity. Overall, the 2SLS estimates corroborate the baseline results, indicating that ESG engagement causally enhances firms’ GIN.

5.5. Propensity Score Matching (PSM)

We employed PSM with nearest neighbor and kernel matching to reduce the selection bias between high- and low-ESG firms. This method ensures that comparisons are made among firms with similar observable characteristics. Firms with ESG scores above the median were assigned to the treatment group, while those with scores below the median comprised the control group. A Probit model including all baseline control variables was used to estimate the probability of high ESGPerf, and propensity scores were calculated for each firm-year. Matching was conducted using a 0.01 caliper. As shown in

Table 8. Propensity score matched regression analysis.

Variables	Nearest Neighbor	Kernel Matching
High ESG (Treatment)	0.0156 *** (3.12)	0.2670 ** (2.39)
ROA	−0.1113 ** (−3.08)	−0.1457 *** (−3.82)
LEVG	−0.2694 * (−1.99)	−0.2159 (−1.56)
SIZ	0.5083 * (1.83)	0.6809 *** (3.17)
Q	0.0028 (0.12)	−0.0094 (−0.49)
LIQ	−0.0014 (−0.02)	−0.0476 (−0.93)
BORDZ	−0.0050 (−0.65)	0.0038 (0.28)
BGDV	−0.0143 * (−1.84)	−0.0011 (−0.13)
GDP	0.0455 *** (5.24)	0.0648 *** (8.58)
INTR	−0.0751 *** (−4.95)	−0.1080 *** (−7.06)
INFL	−0.1262 *** (−5.62)	−0.1531 *** (−6.39)
Constant	4.0743 (0.64)	6.7754 (1.29)
Adjusted R2	0.7367	0.7495
Observations	123	135
Weighted Regression	Yes	Yes

This table displays the results of propensity score matching. 1% (***), 5% (**), and 10% (*) present significance levels. All variables are described in the Appendix A. T statistics are reported in parentheses.

, the treatment group exhibited a significantly higher GIN under nearest neighbor (0.0156, t = 3.12) and kernel matching (0.2670, t = 2.39). The effects of key controls such as ROA, leverage, and firm size remain consistent across specifications. These findings confirm that higher ESG engagement robustly enhances corporate GIN even after accounting for potential selection bias.

5.6. Further Estimation Methods

Table 9. Further estimation methods’ results.

Variable	[23]	Driscoll–Kraay Robust Standard Errors	GLM	Quantile (Median)
ESG	0.01198 *** (5.83)	0.01549 *** (5.99)	0.01198 *** (6.12)	0.00710 ** (2.46)
ROA	−0.04943 (−1.44)	−0.12591 (−1.87)	−0.04943 (−1.51)	−0.02142 (−0.47)
LEVG	−0.2405 *** (−4.45)	−0.2727 *** (−3.41)	−0.24050 ** (−2.67)	−0.17528 ** (−2.38)
SIZ	0.00616 (0.13)	0.43190 * (2.56)	0.00616 (0.13)	0.02367 (0.32)
Q	0.01391 ** (2.97)	0.00224 (0.22)	0.01391 *** (3.12)	0.01016 (1.52)
LIQ	−0.03008 (−0.78)	−0.01463 (−0.23)	−0.03008 (−0.82)	0.01346 (0.26)
BORDSIZ	−0.00286 (−0.16)	−0.00572 (−0.32)	−0.00286 (−0.17)	0.01088 (0.42)
BGDV	0.013098 * (2.11)	−0.01392 (−1.50)	0.01309 ** (2.22)	0.00334 (0.37)
GDP	0.05984 *** (6.62)	0.04440 *** (3.55)	0.05984 *** (6.94)	0.08139 *** (6.51)
INTR	−0.10165 ** (−2.34)	−0.07202 ** (−2.79)	−0.10165 ** (−2.65)	−0.13685 ** (−2.14)
INFL	−0.18497 ** (−2.96)	−0.12760 ** (−2.66)	−0.18497 ** (−2.29)	−0.21791 ** (−2.23)
Constant	8.57385 ** (2.55)	8.92997 (−0.45)	8.57385 ** (2.06)	8.45539 ** (2.17)
Observations	460	460	460	460
F/Wald χ2	22.70 ***	64.22 ***	0.77 ***	125.48 ***
Adjusted R2/Pseudo R2	0.6818	0.7438 (within)	0.1154	0.2488

This table reports empirical results from estimating Equation (1) using different estimation methods. 1% (***), 5% (**), and 10% (*) present significance levels. All variables are described in the Appendix A. T statistics are reported in parentheses.

presents the results of several alternative estimation approaches to ensure the robustness of the baseline findings. Specifically, we implemented fixed-effects regressions with [23]’s and Driscoll–Kraay robust standard errors to account for heteroskedasticity and cross-sectional dependence, Generalized Linear Models (GLM) with heteroskedasticity-consistent covariance matrices to relax normality assumptions, and Quantile Regression to examine heterogeneous effects across the distribution of GIN. Across all specifications, ESGPerf consistently exhibits a positive and statistically significant effect on corporate GIN, confirming that higher ESG engagement enhances environmentally oriented innovative outcomes. The consistency of these results across diverse modeling approaches underscores the robustness of the findings and highlights the central role of ESGPerf in shaping GIN among Saudi-listed firms.

5.7. Difference in Difference Method

We exploit the 2015 adoption of the Paris Agreement as an exogenous shock to examine the causal effect of ESGPerf on firm-level GIN in Saudi Arabia. As the Agreement was internationally determined and beyond firm control, it provides a plausibly exogenous setting to identify changes in corporate sustainability behavior.

Using a difference-in-differences (DiD) framework, we treat 2015 as the event year and classify firms into treatment and control groups based on changes in ESG scores between the pre-event (2013–2014) and post-event (2016–2018) periods. Firms in the top tercile of ESG improvement were assigned to the treatment group, whereas those in the bottom tercile served as controls. The model includes firm- and year-fixed effects, and the Treat_Post interaction is the key coefficient of interest.

Table 10. Difference-in-differences test.

Variables	Coefficient
Treatment	0.1367 *** (3.30)
Post	0.8464 *** (6.33)
Treat_Post	0.1352 ** (2.74)
ROA	−0.0847 ** (−2.64)
LEVG	−0.1045 ** (−2.14)
SIZ	0.2500 ** (2.41)
Q	0.0013 (0.18)
LIQ	−0.0238 (−0.57)
BORDSZ	−0.0016 (−0.14)
BGDV	−0.0013 (−0.26)
GDP	0.0506 *** (12.39)
INTR	−0.0835 *** (−11.08)
INFL	−0.0776 *** (−7.74)
Constant	1.880 ** (2.83)
Observations	112
Within R2	0.8951
Wald χ2	366.85 ***

This table displays the results of Difference-in-differences test for the relationship between ESGPerf and GIN. 1% (***) and 5% (**) present significance levels. All variables are described in the Appendix A. T statistics are reported in parentheses.

shows the regression results. The Treat_Post coefficient was positive and statistically significant (0.1352, t = 2.74), indicating that firms with greater ESG improvements experienced a larger increase in GIN after 2015. The Treatment (0.1367 ***) and Post (0.8464 ***) coefficients are also positive and significant, suggesting higher baseline innovation among treated firms and a general post-agreement increase in GIN.

Overall, the results in Table 10 provide robust evidence that improvements in ESGPerf, reinforced by the Paris Agreement, significantly enhanced GIN among Saudi firms.

5.8. Heterogeneity Analysis

To investigate whether the impact of ESGPerf on GIN varies across firms, we conduct a heterogeneity analysis focusing on industry pollution level and firm size. This approach allows us to explore whether firm-level environmental exposure and organizational scale moderate the effectiveness of ESGPerf in promoting GIN. Specifically, firms operating in polluting industries may face higher regulatory and climate-related risks, which could influence their innovation behavior differently from non-polluting firms. Similarly, large and small firms may differ in their capacity, responsiveness, and strategic orientation toward ESG-driven innovation. By examining these dimensions, we aim to uncover the nuanced ways in which ESGPerf interacts with firm characteristics to shape GIN outcomes in emerging markets.

5.8.1. Industry Pollution Level

The results of the heterogeneity analysis based on industry-level pollution are shown in

Table 11. Influence of ESG on GIN: Polluting vs. non-Polluting firms.

Variables	Polluting Firms	Non-Polluting Firms
ESG	0.0216 * (2.12)	0.0156 ** (2.27)
ROA	0.1373 * (1.96)	0.4396 *** (11.14)
LEVG	−0.2248 (−0.49)	−0.0872 (−0.51)
SIZ	0.9422 (1.43)	0.1191 (0.41)
Q	−0.0104 (−0.20)	−0.0179 (−0.54)
LQD	0.0839 (0.51)	0.2385 *** (7.18)
BRDZ	−0.0003 (−0.01)	−0.0007 (−0.07)
BGDV	−0.0314 (−1.12)	−0.0164 (−1.39)
Constant	0.353 (0.75)	0.3947 (0.06)
Macroeconomic controls	Included	Included
Firm FE	Yes	Yes
Year FE	Yes	Yes
Observations	154	306
Within R2	0.7815	0.6722
F-statistic	12.97 *	776.36 ***
Sigma_u	1.846	1.095
Sigma_e	0.338	0.262
Rho	0.968	0.946

This table displays the empirical results of estimating the impact of ESG on GIN for polluting (Energy, oil, basic materials and utilities) and non-polluting (others) firms. 1% (***), 5% (**), and 10% (*) represent significance levels. All variables are described in the Appendix A. The T-statistics are reported in parentheses.

. We reveal notable differences in the ESGPerf-GIN relationship between polluting and non-polluting firms. For polluting firms, ESGPerf positively influences GIN (coef = 0.0216, t = 2.12), though the effect is marginally significant, suggesting that these firms primarily engage in GIN as a risk-mitigation response to regulatory and climate pressures. In contrast, non-polluting firms exhibit a significant and positive ESGPerf-GIN association (coef = 0.0156, t = 2.27), indicating that these firms adopt GIN proactively, leveraging ESGPerf to achieve strategic sustainability goals rather than merely complying with environmental requirements. Financial capacity further amplifies this effect for non-polluting firms, as liquidity is positively associated with GIN, whereas profitability appears to constrain investment in green technologies for both firm types. Overall, these findings demonstrate that firm-level environmental exposure and climate risk shape the effectiveness of ESG initiatives, highlighting the need for tailored ESG strategies in emerging markets to stimulate GIN.

5.8.2. Firm Size

The heterogeneity analysis by firm size reveals that the effect of ESGPerf on GIN differs markedly between large and small firms. As shown in

Table 12. Influence of ESG on GIN: Large vs. small Firms.

Variables	Large Firms	Small Firms
ESG	0.0078 * (1.81)	0.0253 *** (5.77)
ROA	−0.1127 * (−1.93)	−0.3453 *** (−6.35)
LEVG	0.0184 (−0.12)	−0.2819 * (−1.89)
SIZ	1.2139 ** (2.61)	0.2316 (0.85)
Q	−0.0548 *** (−3.49)	−0.0174 (−0.79)
LQD	0.1235 (1.16)	0.1731 *** (5.18)
BRDZ	0.0353 (−1.02)	−0.0148 (−0.30)
BGDV	−0.0185 (−1.49)	−0.4472 ** (−3.67)
Constant	−19.453 (−1.56)	−0.9655 (−0.14)
Macroeconomic controls	Included	Included
Firm FE	Yes	Yes
Year FE	Yes	Yes
Observations	284	176
Within R2	0.703	0.8247
F-statistic	140.34 ***	20.09 *
Sigma_u	1.343	2.01
Sigma_e	0.264	0.31
Rho	0.963	0.977

This table displays the empirical results of estimating the impact of ESG on GIN for large (above the average sample size) and small (other) firms. 1% (***), 5% (**), and 10% (*) represent significance levels. All variables are described in the Appendix A. The T-statistics are reported in parentheses.

, for large firms, ESGPerf has a marginally positive impact (coef = 0.0078, t = 1.81), indicating that while ESG initiatives support GIN, the effect is relatively modest. This may reflect the fact that large firms already possess established R&D capabilities and structured innovation pipelines, so incremental ESG improvements have a smaller marginal influence. In contrast, small firms exhibit a strong and significant relationship (coef = 0.0253, t = 5.77), suggesting that ESGPerf is a key driver of GIN for firms with fewer resources and less institutional inertia. Financial factors further reinforce this pattern: liquidity significantly promotes GIN in small firms, whereas firm size enhances innovation in large firms, consistent with their organizational scale advantages. Overall, these findings suggest that firm size moderates the ESGPerf-GIN relationship, with ESG policies more strongly incentivizing proactive innovation in smaller firms, while larger firms respond more steadily, reflecting structural and resource-related constraints.

6. Conclusions

This study evaluates the impact of ESGPerf on GIN among Saudi-listed firms in 2002–2024, with climate physical risk as a moderating variable. The evidence using the Fixed Effects Model confirms that overall ESG engagement positively affects GIN, with Social and Environmental dimensions as the most influential pillars. This underscores that companies that take a proactive approach to environmental and social responsibilities are better positioned to develop novel green technologies and processes, thereby enhancing their competitiveness and long-term resilience.

Moderation analysis further demonstrates that increased climate risk, measured by the CPRI, weakens the positive relationship between ESGPerf and GIN. This finding highlights that climate uncertainty not only challenges ESG effectiveness but also encourages firms to adopt adaptive innovation strategies. That is, environmental uncertainty assessed by climate risk stimulates firms to adopt adaptive innovation strategies, reflective of proactive adaptation in the face of volatile external environments. Furthermore, robustness tests based on various estimation techniques and alternative measures of variables confirm the stability and reliability of these findings and lend assurance to the research.

This study contributes to the literature by providing firm-level evidence from the Saudi Arabian setting, demonstrating the critical role of ESG—particularly its Social and Environmental pillars—in driving GIN in a resource-based economy. It also highlights the moderating role of climate risk, showing that while such risk can undermine ESG effectiveness, it reinforces the importance of climate-aware strategic adaptation. These findings offer practical implications for Saudi policymakers and business leaders by emphasizing the need to integrate ESG practices with climate risk management to support sustainable innovation and align with Vision 2030 objectives.

Despite the contributions of this study, some limitations have been identified. First, this study only considers Saudi-listed firms. This may limit the generalizability of the results to other emerging and developed countries. Second, this study used the CPRI as the sole measure of climate risk, which may not capture transitional and policy-related environmental uncertainties. Additionally, only quantitative ESG scores and GIN outputs are analyzed, while qualitative aspects of innovation may reveal deeper mechanisms.

Future research could extend this analysis to other emerging and developed economies to determine the generalization of the ESGPerf-GIN relationship in diverse institutional settings. Furthermore, the study may be extended to other types of climate risks that account for transition and policy risks in a more comprehensive analysis of the role of environmental risks in GIN. The study may also be extended to explore the qualitative aspects of GIN in resource-based firms in Saudi Arabia to identify the processes through which ESG commitments contribute to GIN. Moreover, examining the dynamic and real-time interactions between ESG practices, climate risk, and innovation would be particularly valuable given the rapidly evolving technological and regulatory environments.

Overall, the findings demonstrate that the effectiveness of ESG engagement in driving GIN is contingent on climate risk, underscoring the need to align sustainability strategies with risk management in uncertain environments.