From risk to resilience: how ESG uncertainty drives R&D investment

Despite ongoing debate about the true meaning and financial impact of implementing environmental, social and governance (ESG) factors, they have gained substantial attention from both academics and practitioners due to their potential influence on corporate decision-making and financial performance. Companies that proactively integrate ESG considerations into their strategies may experience enhanced financial performance, improved reputation and increased competitiveness. For instance, a meta-analysis by Friede, Busch, and Bassen (2015) found that approximately 90% of studies reported a non-negative relationship between ESG factors and corporate financial performance. However, the lack of standardized ESG reporting frameworks poses significant challenges. Berg, Kölbel, and Rigobon (2022) noted that only about 60% of ESG ratings are concordant, compared to 99% for credit ratings from major agencies. Moreover, the potential for “greenwashing,” where companies falsely portray their practices as environmentally friendly, remains a concern. These mixed outcomes reflect the complex environment firms face as they navigate the growing pressure to integrate ESG considerations into their strategic planning, not only for ethical reasons but also to remain competitive and align with stakeholder expectations.

Alongside the growing focus on ESG, the uncertainty surrounding ESG-related policies and regulatory changes presents opportunities and challenges for firms navigating this evolving landscape. Firms must manage the complexities of regulatory compliance, reporting standards and stakeholder expectations to address these challenges and capitalize on emerging opportunities. In this paper, we examine the impact of ESG policy uncertainty on firms' innovation, proxied by R&D investment. Specifically, we use the ESG-Related Uncertainty Index (ESGUI) developed by Ongan, Gocer, and Işık (2025), which applies text analysis techniques to measure the frequency of ESG-related and uncertainty-related terms in economic reports. Given the importance of R&D investment as a driver of innovation and a competitive advantage, understanding the link between ESG uncertainty and corporate innovation strategies is crucial.

We find that ESGUI is positively correlated with R&D investment, suggesting that higher ESG-related uncertainty leads to increased R&D investment. This holds across both industry and firm fixed-effect models. High cost of goods sold (COGS) may amplify the need for firms to engage in innovation when ESG uncertainty is high. We next provide a channel test to examine this intuition. The result reveals that firms with higher production costs are more likely to increase R&D investment under heightened ESG uncertainty. This supports the theory that cost-reduction incentives drive innovation investment in response to ESG uncertainty.

We also examine the moderating role of firm ownership structure. State-owned enterprises (SOEs) often benefit from government support and policy stability, which can reduce the need to engage in risk-mitigation strategies such as innovation. To test this, we included an interaction term between ESGUI and SOE status in our regressions. The results indicate a significantly negative coefficient in the interaction term, suggesting that SOEs are less responsive to ESG uncertainty in terms of R&D investment. This implies that government-backed stability can dampen the pressure for innovation in uncertain ESG contexts, highlighting the importance of institutional and ownership factors in shaping corporate responses to ESG risks.

In addition, we explore the moderating role of firms' ESG performance. We find that firms with higher ESG scores exhibit a significantly stronger relationship between ESGUI and R&D investment, suggesting that companies already positioned as ESG leaders face greater reputational pressures and are more proactive in leveraging innovation to manage ESG-related risks.

To further address potential endogeneity concerns, we employ an instrumental variable (IV) approach using changes in China's particulate matter 2.5 (PM2.5) concentration as an instrument for ESGUI. The rationale is that nationwide fluctuations in air pollution influence environmental and sustainability concerns, thereby affecting ESG uncertainty at the country level, while having no direct effect on individual firms' R&D decisions once ESG uncertainty is accounted for. The two-stage least squares (2SLS) results confirm that ESGUI has a positive and significant impact on R&D investment, providing stronger causal evidence that firms increase innovation in response to ESG-related uncertainty. These findings reinforce our main results and strengthen the basis for the theoretical and policy implications discussed in this paper.

To verify that the main findings are not sensitive to specific measurement choices, we also provide evidence on robustness tests by adopting alternative measures in these analyses. We scale R&D investment by operating income instead of total assets as an alternative measure of R&D. Besides, we also employ an alternative metric of the China-specific ESGUI, the global ESGUI. We find that ESGUI remains significantly positive in all estimations, confirming that the core findings hold under different measurement approaches.

Finally, to determine whether ESG uncertainty has a lasting impact on firms' innovation strategies beyond the immediate period, we include lagged ESGUI values in the regression model. The empirical analysis strongly supports the notion that ESG-related uncertainty drives firms to increase R&D investment as a risk mitigation strategy. These findings are robust across different measurement approaches and persist over time, suggesting that firms strategically adapt to ESG uncertainty by prioritizing innovation.

Our study contributes to three strands of literature. First, we contribute to an emerging literature on ESG uncertainty. The concept of policy uncertainty affecting firm behavior has been widely studied in economics and finance. Prior research has established that regulatory uncertainty impacts corporate investment decisions, as firms tend to be more cautious in uncertain environments (Ahir, Bloom, & Furceri, 2022; Huang & Yang, 2025). Similar patterns emerge in ESG-related contexts, where firms face increasing pressure to adopt sustainable practices yet the evolving regulatory landscape creates uncertainty in strategic decision-making (Xie, Ali, Kumar, Naz, & Ahmed, 2024). Beyond R&D, ESG uncertainty has been linked to broader corporate outcomes, including financial performance, market valuation and risk management strategies (Zhang & Wang, 2020). Firms that proactively engage with ESG challenges tend to develop competitive advantages, whereas those that remain passive face higher operational risks (Chung, Dai, & Elliott, 2022). This highlights the strategic importance of ESG-related decision-making, particularly in industries that are highly exposed to regulatory shifts and investor scrutiny.

This paper also enriches the literature on factors that affect financial decisions and R&D investment. R&D investment plays a crucial role in corporate strategy, innovation and long-term growth. Firms allocate resources to R&D activities to gain competitive advantages, improve efficiency and respond to technological and regulatory changes (Hall & Lerner, 2010). The primary determinants of R&D investment are the availability of financial resources (Aghion, Van Reenen, & Zingales, 2004; Brown, Martinsson, & Petersen, 2009; Czarnitzki & Hottenrott, 2011; Hall & Lerner, 2010), firm size, ownership structure, corporate governance (Baysinger, Kosnik, & Turk, 1991; Cohen & Levinthal, 1990; Li & Xia, 2018), market competition and industry dynamics (Aghion, Bloom, Blundell, Griffith, & Howitt, 2005; Schumpeter, 1942), government policies and support (David, Hall, & Toole, 2000; González & Pazó, 2008; Xie et al., 2024), and external collaborations and knowledge spillovers (Feldman & Florida, 1994; Griliches, 1992; Hagedoorn, Link, & Vonortas, 2000). Our study provides evidence on the impact of ESG-related uncertainty on corporate R&D investment.

Our research also contributes to the literature on the role that state ownership plays. We not only provide empirical evidence on the relationship between ESG uncertainty and firms' innovation activities but also investigate the moderating role of state ownership (SOE) in this relationship, recognizing that government-backed firms may exhibit distinct responses to ESG-related uncertainty due to their financial stability and policy support. The findings shed light on how firms navigate uncertainty-induced innovation incentives, with implications for policymakers, investors and corporate managers.

In sum, our research offers valuable insights for stakeholders and policymakers. A solid comprehension of the dynamics between trade policy uncertainty and a firm's strategic decisions regarding liquidity and investments is crucial for companies to manage risks, make informed plans and maintain a competitive advantage.

The remainder of the paper is organized as follows. We discuss the literature and hypotheses in Section 2. Section 3 outlines the data, sample formation and methodology. Section 4 presents the main empirical results, explores the underlying mechanisms and moderating factors and reports the instrumental variable analysis along with robustness checks. Section 5 concludes the paper with final remarks.

Uncertainty surrounding ESG policies often affects corporate investment strategies but its effects can be mixed and two-sided. On the one hand, ESG uncertainty can serve as a key driver of a firm's innovation, promoting long-term growth and competitiveness. Prior studies suggest that firms adjust their innovation activities in response to economic and regulatory uncertainty (Thatcher & Oliver, 2001). From a resource-based view (Barney, 1991), ESG uncertainty motivates firms to develop distinctive and hard-to-imitate capabilities – such as sustainable technologies – that strengthen their competitive advantage. Similarly, drawing from stakeholder theory (Freeman & Phillips, 2002), firms may increase investment under ESG uncertainty to address the expectations of critical stakeholders, including investors and regulators and to maintain access to essential resources. The rising demand for sustainable innovations driven by ESG concerns has further reinforced these strategic behaviors. Empirical evidence also indicates that firms facing higher ESG uncertainty tend to invest more in innovation to mitigate risks and exploit emerging opportunities (Dang, Nguyen, Lee, Nguyen, & Le, 2023). However, the magnitude of such investment responses depends on firm-specific characteristics, industry dynamics and external market conditions.

On the other hand, under the real options theory (Dixit, 1994), firms may postpone or scale back investment to preserve flexibility when faced with high uncertainty. Accordingly, ESG-related regulatory ambiguity can lead firms to delay long-term investment projects or reallocate capital toward less ESG-sensitive assets. Similarly, high policy uncertainty regarding carbon emissions regulations or sustainability disclosures can discourage firms from undertaking irreversible capital expenditures in green technologies (Bolton & Kacperczyk, 2021). From an institutional theory perspective, changes in ESG policies create institutional pressures that increase compliance risks, prompting firms to adopt conservative investment behavior. Empirical studies support these arguments: Gibson et al. (2022) found that ESG policy shifts create investment distortions, leading firms to increase cash holdings and reduce capital expenditures in the short run. Kacperczyk, Sundaresan, and Wang (2021) demonstrate that firms in industries with high ESG regulatory uncertainty experience a higher cost of capital, reducing investment efficiency. Moreover, precautionary motives suggest that firms operating in countries with stringent or unpredictable ESG policies tend to favor flexibility by holding liquid assets instead of committing to ESG-focused projects (Dyck, Lins, Roth, & Wagner, 2019).

Based on the above arguments, we propose two competing hypotheses:

Our key variables are derived from various sources covering 2003 to 2023. The primary independent variable, the ESGUI, was developed by Ongan et al. (2025) using text analysis techniques applied to monthly country reports from the Economist Intelligence Unit (EIU). These reports provide comprehensive economic, political and sector-specific information for developed and developing countries (Xie et al., 2024). The ESGUI, covering 25 countries, is constructed using a three-step process. First, the environmental (E), social (S) and governance (G) sub-indices are formed by extracting ESG-related keywords from EIU reports, calculating their relative frequencies to the total word count using natural language processing (NLP) techniques and normalizing the results using the Min-Max Scaler. Second, an uncertainty sub-index is developed using the same methodology but focusing on uncertainty-related keywords, following the World Uncertainty Index (WUI) framework by Ahir et al. (2022). Finally, the ESGUI is derived for each country by averaging the ESG and uncertainty sub-indices on a monthly basis, with a global ESGUI constructed by aggregating country-level indices. Following Chung et al. (2022) and Dang et al. (2023), the index values are normalized to range from 0 to 100. It is worth noting that this ESGUI measure has been widely used in recent literature across a range of studies (Cui & Maghyereh, 2025; Gaies, 2025; Olanrewaju, Adebayo, & Uzun, 2025).

Since our sample focuses on listed firms in China, we use the China ESGUI, which specifically captures uncertainty shocks in the Chinese domestic context. In contrast, the Global ESGUI is constructed by aggregating country-level indices across 25 nations to represent broad international ESG sentiment; we utilize this global measure as an alternative in our robustness tests. To convert the monthly ESGUI data into an annual measure, we compute the yearly average of the monthly values and apply a log transformation.

The firm-level data are obtained from the China Stock Market and Accounting Research database. We scale R&D investment by total assets and, in robustness tests, alternatively by operating income. Our regression includes the following control variables: Size, as larger firms often have more resources for R&D (Becker, Hottenrott, & Mukherjee, 2022); Return on Assets (ROA), which indicates the efficiency of asset use in generating earnings and influences R&D investment decisions (Rađenović et al., 2023); Tangibility, measured as the ratio of tangible to total assets, as firms with more tangible assets may exhibit different investment behaviors due to greater stability and easier access to financing; Leverage, since a firm's debt level impacts its risk profile and investment capacity (Gebauer, Setzer, & Westphal, 2018); and Cash Holding, representing the liquid assets available to fund R&D activities without relying on external financing (Denis et al., 2010).

We begin our analysis by examining the influence of ESGUI on firms’ R&D investment. Specifically, we conduct the following ordinary least squares (OLS) regression model.

where Y denotes our outcome variable of interest, $Xit$ is a vector of control variables and $αi$ represents firm fixed effect ^[1]. To ensure data integrity, we implement several exclusion criteria: firm-year observations with incomplete data; companies in finance and utility sectors; and firms undergoing special treatments (ST) ^[2]. In addition, our continuous variables are winsorized at the 1% and 99% levels to mitigate the influence of outliers in the dataset. Variable definitions are presented in Table 1, and summary statistics of all key variables are reported in Table 2.

We first investigate the impact of the ESGUI on firms' R&D investments. Across all columns in Table 3, the coefficients of ESGUI are significantly positive, indicating that higher ESGUI levels are associated with a significant increase in a firm's R&D investment. The magnitude is also economically meaningful. For example, in Table 3, Column 3, a one-standard-deviation increase in ESGUI is associated with a 0.8% rise in R&D investment (0.3523 × 0.0224), corresponding to a 75% increase relative to the sample mean (0.3523 × 0.0224/0.0105). This suggests that when firms face greater uncertainty related to ESG factors, they respond by increasing their commitment to innovation and long-term strategic projects. One possible explanation is that heightened ESG-related uncertainty may prompt firms to invest more heavily in R&D to enhance their adaptability and resilience in a rapidly changing business environment. The consistency of the results across these models reinforces the reliability of the observed relationship and suggests that the positive link between ESGUI and R&D investment is not driven by industry- or firm-specific characteristics. These results highlight the strategic importance of R&D investment when firms navigate ESG-related uncertainty and maintain a competitive edge.

The existing literature widely documents that innovation strengthens a firm's competitive advantage amid external uncertainties by enhancing operational efficiency and productivity (Thatcher & Oliver, 2001) and by reducing costs (DeGraba & Sullivan, 1995). We posit that high production costs, measured by the COGS, can serve as a channel linking ESG uncertainty and innovation. Rather than assuming a direct mechanical effect, we interpret COGS as capturing firms' exposure to cost pressures, which may vary systematically across firms and sectors. During periods of heightened ESG uncertainty, firms, particularly those with higher baseline cost structures, may be more sensitive to changes in input costs, compliance requirements, or supply chain conditions driven by the evolving ESG context. To the extent that ESG-related developments amplify such cost pressures, firms with higher COGS may have stronger incentives to seek efficiency gains. These pressures can place additional financial strain on firms (Alfaro, Bloom, & Lin, 2018; Chava, Oettl, Subramanian, & Subramanian, 2022; Chen, Dong, & Lin, 2020).

These dynamics are consistent with the Porter Hypothesis (Porter & Linde, 1995), which suggests that rising environmental or social pressures, while potentially increasing compliance and production costs, can stimulate innovation that enhances competitiveness. Similarly, drawing on the dynamic capabilities framework (Teece, Pisano, & Shuen, 1997), firms may respond to such pressures by investing in R&D to reconfigure resources, improve operational efficiency and strengthen adaptive capacity. Importantly, this mechanism does not imply that all high-COGS firms are inherently more innovative; rather, it suggests that, conditional on facing ESG-related uncertainty, firms under greater cost pressure may have stronger incentives to innovate. Thus, we posit that higher production costs can incentivize innovation as a strategic response to improve efficiency, reduce long-term risks and enhance organizational resilience.

Table 4 presents our regression estimates. In Column 1, we find that ESGUI significantly increases COGS. In Column 2, both COGS and ESGUI are positively and significantly associated with R&D investment. These results suggest that higher ESG uncertainty leads to higher costs, which in turn drive greater R&D investment. Together, these findings provide empirical support for the notion that ESG uncertainty influences corporate R&D through cost-related channels.

Next, we examine the moderating role of state-owned enterprises (SOEs) in the relationship between ESG uncertainty (ESGUI) and R&D investment. In China's institutional environment, state ownership allows the government to exert substantial influence over corporate strategy, decision-making and resource allocation (Huang, Tarkom, & Yang, 2024; Huang & Yang, 2024a; 2024b). This government involvement ensures that SOEs align closely with national interests and economic priorities, granting them advantages such as preferential access to financing, policy support and greater resilience to market fluctuations (Dong, Hou, & Ni, 2021; Jin, Wang, & Zhang, 2023; Zhang & Wang, 2020). Moreover, SOEs benefit from implicit government guarantees, often receiving financial assistance or favorable policy treatment during economic downturns, which reduces their exposure to market volatility and uncertainty. As a result, while private firms may respond to ESG-related uncertainty by increasing R&D investment to enhance competitiveness and efficiency, SOEs face weaker incentives to engage in riskier innovation activities. Their government-backed stability and access to resources mitigate the financial and operational pressures that typically drive innovation under uncertainty. Consequently, empirical evidence suggests that SOEs are less responsive to uncertainty-driven innovation incentives, as their risk-mitigation strategies rely more on state support than on proactive R&D investment (Dong et al., 2021; Huang et al., 2024).

As shown in Table 5, Column 3, the coefficient of the interaction term between ESGUI and SOE is significantly negative. Combined with the positive effect of ESGUI on R&D investment, this finding suggests that while ESG uncertainty generally encourages firms to invest in innovation, SOEs are less responsive to this incentive due to their government-backed stability. This empirical evidence supports the theory that the financial and strategic advantages enjoyed by SOEs reduce the need for R&D investment as a risk-mitigation strategy under heightened ESG uncertainty.

The impact of macro-level ESG uncertainty on corporate policies is likely to be contingent on a firm's actual ESG performance. Accordingly, we examine how ESG performance moderates the relationship between ESGUI and R&D investment. Our results are reported in Table 6. The findings in this table indicate that ESG performance significantly strengthens this relationship, with the ESGUI*ESG interaction being positive and economically meaningful.

We interpret this effect through a reputation-preservation lens: when the direction of ESG tightening is clear but the timing and design remain uncertain, high-ESG firms face greater reputational risk from any perceived backsliding. Markets punish corporate social responsibility (CSR)-related missteps more severely for firms with high visibility and customer awareness (Krüger, 2015; Servaes & Tamayo, 2013). Because stakeholder expectations are higher for these firms, lapses that might be tolerated for an average peer trigger sharper scrutiny (Krüger, 2015). In this context, R&D acts as a credible, capability-building investment that strengthens absorptive capacity and organizational learning, while also sending a more meaningful signal of preparedness than symbolic CSR gestures (Cohen & Levinthal, 1990; Zahra & George, 2002). Furthermore, heightened uncertainty increases the value of flexible, real-option-like investments. Theory and evidence suggest that under such conditions, exploratory R&D becomes more valuable, expanding the firm's range of options for future regulatory and market scenarios (Dixit & Pindyck, 1994; Trigeorgis & Reuer, 2017).

To address potential endogeneity between firms' R&D investment and ESG uncertainty (ESGUI), we employ an instrumental variables (IV) approach, using changes in China's PM2.5 concentration as the instrument. The rationale is that nationwide fluctuations in air pollution heighten environmental and sustainability concerns, thereby increasing operational uncertainty for firms and amplifying ESG-related uncertainty at the national level (Guo, Pan, & Zhao, 2025). However, once ESG uncertainty is accounted for, aggregate PM2.5 changes are unlikely to directly affect individual firms' R&D spending, making this a relevant and plausibly exogenous instrument.

Table 7 presents the results of our 2SLS estimations. The first-stage regression shows that changes in China's PM2.5 concentrations significantly increase ESG uncertainty, confirming strong instrument relevance. Untabulated statistical diagnostics further validated the robustness of our IV approach. The Kleibergen–Paap rk Lagrange Multiplier statistic (χ² = 2567.65) strongly rejects the null of under-identification, while the Cragg–Donald Wald F statistic (3630.39) far exceeds conventional thresholds, ruling out weak instrument concerns. The Anderson–Rubin Wald test (F = 13.45; χ² = 13.45) and the Stock–Wright LM S statistic (χ² = 14.28) both confirm that the instrument is powerful and relevant, and that inference remains valid even under potential weak identification. Finally, the Hansen J statistic (p = 0.000) indicates that the overidentifying restrictions are satisfied, supporting the exogeneity of the instrument.

In the second stage, we find that ESG uncertainty exerts a positive and statistically significant effect on firms' R&D investment. This result suggests that heightened ESG-related uncertainty prompts firms to allocate more resources to innovation, likely as a strategic response to anticipated regulatory shifts, stakeholder expectations and long-term sustainability risks. We note, however, that the instrumental variable is more closely related to the environmental dimension of ESG and we therefore acknowledge that it may not fully capture variation in the social or governance components. We interpret our results with appropriate caution. Nonetheless, environmental shocks are an important driver of broader ESG-related uncertainty, as they often trigger regulatory, reputational and stakeholder responses across multiple dimensions.

Overall, the evidence from this analysis provides strong support for the validity of using China-level PM2.5 changes as an instrument for ESG uncertainty. The 2SLS estimates reveal a robust causal link: increases in ESG uncertainty lead firms to expand their R&D investments. These findings underscore ESG-related uncertainty as a catalyst for corporate innovation and highlight firms' proactive adaptation to heightened sustainability challenges.

To ensure the robustness of our baseline findings, we employ alternative measures of both R&D investment and ESGUI. In Columns 1 to 3 of Table 8, we scale R&D investment by operating income instead of total assets. The results remain consistent, showing a significant positive impact of ESGUI on firms' R&D investment, reinforcing our primary conclusions. In Columns 4 to 6, we further validate the credibility of our findings by replacing the China-based ESGUI with the Global ESGUI. Once again, we find a consistently significant positive effect of ESGUI on firms' innovative activities. These results confirm the reliability and generalizability of our main findings across different measurement approaches.

In Table 9, we conduct an in-depth analysis of the lagged effect of ESGUI on firms' R&D investment to determine whether the influence of ESG uncertainty extends beyond the current period. The results reveal that the coefficients of lagged ESGUI remain significantly positive across all three columns, suggesting that firms continue to increase their innovation investment in response to past ESG uncertainty. This pattern is consistently observed across different estimation models, reinforcing the validity of our findings. These results indicate that the impact of ESG uncertainty on R&D investment is not only immediate but also sustained over time, highlighting the long-term strategic adjustments firms make in response to ESG-related risks and uncertainties.

This study examines the impact of ESG-related uncertainty on corporate R&D investment, focusing on the moderating role of state ownership. Using a novel ESGUI for publicly listed firms in China, our empirical analysis shows that heightened ESG uncertainty is associated with significantly higher R&D investment. This finding is consistent with the view that firms may respond to regulatory and policy ambiguity by strengthening their innovative efforts to mitigate potential risks and enhance long-term resilience. Our findings also highlight the role of production costs as a key channel through which ESG uncertainty affects R&D investment. Firms with higher production costs are more sensitive to ESG uncertainty and are more likely to invest in innovation as a cost-reduction strategy.

These findings carry important implications for both corporate leaders and policymakers. From a managerial perspective, our results suggest that ESG-related uncertainty need not be purely constraining; rather, it can catalyze innovation. Firms, particularly those facing higher production costs, may benefit from proactively adjusting their R&D strategies. Such strategies may involve prioritizing efficiency-enhancing innovations and adopting flexible investment approaches to address evolving ESG risks. From a policy standpoint, our evidence indicates that ESG uncertainty has real effects on firms' innovation behavior. The weaker responsiveness of state-owned enterprises highlights the role of policy stability and government support, while the stronger response among high-ESG firms suggests complementarities between existing ESG capabilities and innovation incentives. These findings underscore the importance of designing ESG policies that balance clarity and flexibility, thereby reducing uncertainty while still preserving firms' motivation to innovate.

This paper is not without limitations. First, given that our instrumental variable, while relevant, is more closely related to the environmental dimension of ESG, future scholars can explore alternative instruments that better capture the full scope of ESG variation. Second, ESG-related uncertainty is measured using a composite index, which does not allow for a disaggregated analysis across its underlying dimensions. Future research could build on this work by examining these components separately to provide more granular insights.