Competing Institutional Logics in Corporate ESG: Evidence From Developing Countries

2.1 Competing Logics in Corporate Sustainability

An organization is formed by entities with shared ethical objectives. However, they may possess different logics stemming from distinct beliefs, values, and axioms through which individuals define the meaning of life (Thornton et al. 2012). As the construction of meaning varies across different agents, the dominant institutional orders, including market, community, state, corporation, religion, profession, and family, determine the objectives of institutions (Reay and Hinings 2009), which is eventually transmitted into their decision-making process and strategies (Almandoz 2014). For the sake of brevity, our analysis focuses on market and community logics, consistent with Venkataraman et al. (2016) and Chu et al. (2024) (see Table 1).

The market logic is based on the assumption of a free and competitive market, where the economic agents are utility maximization entities with self-interest norms (DiMaggio and Powell 1983). Its main authority comes from shareholders' activism driven by their expectations of the firm value, making operational efficiency and profit the heart of its business strategies (Skelcher and Smith 2015). Therefore, as far as the corporate ESG is concerned, this institutional order considers value as a mare financial sustainability.

On the other hand, the community logic champions equity and common (social) interests (Thornton et al. 2012). It represents collective connections among individuals, emphasizing interpersonal and personalized relationships. Customers are among the most important communities that have been growingly demanding more sustainable products, as consistently suggested by recent market studies (Bar Am et al. 2023). They endorse the attainment of intergenerational, intragenerational, and physiocentric justice. This institutional order is also embodied in the ESG-related nongovernmental organizations (NGOs) and lobbyists.

2.2 The Dominance of Market Logic

Indeed, distinct institutional logics are evident not only between investors and managers but also among various types of investors. Starks (2023) divides investors into four different types: (i) traditional, (ii) classic ESG, (iii) socially responsible, and (iv) impact investors. Traditional investors follow a market-dominant logic, aiming solely to maximize risk-adjusted returns. Classic ESG investors, while also return-focused, limit their investment universe by excluding “sin sectors” to maintain ESG compliance. However, their investment strategy remains to obtain a greater-than-market return. Socially responsible investors are willing to accept slightly lower-than-market returns to achieve social and environmental impact. Although market logic still influences them, their focus shifts toward community by maximizing integrated values that combine financial, environmental, and social outcomes. Finally, impact investors, which represent a small minority in the market, prioritize environmental and social outcomes and accept lower risk-adjusted returns, as their dominant logic centers on community logic (see Appendix A for a summary of the spectrum of investors' logic).

A plethora of empirical literature suggests that the proportion of impact investors remains small. Giglio et al. (2023) document that sustainable investors who consider ethics as their motive for investments are only 25%, whereas other respondents consider risk, return, and other pecuniary benefits as their primary motives. The results are derived from US-based investors who already have relatively higher awareness regarding ESG as compared to developing economies (UNGC 2023). Krueger et al. (2020) also point out that 5 out of 7 highest motives in ESG investing are influenced either by market or state, namely, return, risk, tail risk, reputation, and legal. Riedl and Smeets (2017) argue that social preference is not the only motive of ESG investing; investors also engage with sustainable assets for social signaling purposes. They also find that the social preference motive increases the likelihood of investors choosing ESG corporations by 14% but does not influence the amount of money they allocate to such an investment. Giglio et al. (2023) confirm Riedl and Smeets (2017) that, agnostic to the motives, investors only allocate meaningful amounts of investments on ESG should they anticipate the investment to surpass market performance.

Given that a significant portion of investors are either traditional or classic ESG, it is plausible to expect market logic to dominate investor behavior, leading to adverse reactions toward firms' engagement in ESG activities. Contrary to the mainstream empirical studies, this study supports the negative relationship between firms' ESG performance and their market values, consistent with Di Giuli and Kostovetsky (2014) and Behl et al. (2022). This relationship is even more pronounced in the spatial context of developing economies, where investors' internal motivation toward social and impact investments is more limited (UNGC 2023). In such contexts, investors often perceive the firm's adherence to ESG principles as non-revenue-generating, and therefore, they penalize its expected future cash flow, resulting in a lower market value.

Unlike shareholder theory, which suggests that firms engaging in ESG activities may suffer financially, institutional theory predicts that investors' negative sentiment toward ESG may stem from preconceived assumptions or beliefs rather than from the actual impact of ESG on firm performance. As Gümüşay et al. (2024, 5) state, “the propensity to prioritize specific logics impacts how they attend to institutional expectations, demands, and prescriptions.” This explains why negative perceptions persist, even though numerous studies indicate that high-performing ESG firms can achieve superior financial performance compared to, or at least on par with, their low-performing counterparts (Xie et al. 2019; Qureshi et al. 2021; Shin et al. 2023).

There is a surprisingly limited body of empirical literature that applies institutional logic to ESG practices in developing economies. However, related studies have been conducted in the areas of corporate social responsibility (CSR) and religious-based hybrid organizations. Chu et al. (2024) report that market logic remains dominant in the CSR activities of multinational enterprises (MNEs) in India, as these organizations prioritize activities that generate positive externalities over negative ones to enhance their competitive advantages. Similarly, Zhao and Lounsbury (2016) highlight that market logic is not only dominant but also crucial for religious-based microfinance institutions in developing economies to secure larger capital bases, ensuring financial sustainability over outreach and social impact.

Asutay et al. (2023) observe that market logic also prevails in the trading behavior of retail investors in the Islamic stock market in Indonesia, overshadowing religious logic. This is despite the fact that incorporating stricter religious logic into equity investing does not compromise portfolio performance (Alotaibi et al. 2022). Further, Asutay and Yilmaz (2025) expand on this issue in the broader context of emerging economies. They contend that the hegemony of market logic over religious logic is evident in the disembedded nature of Islamic finance practices, where debt-based instruments dominate over equity-based instruments, which are more closely aligned with Islamic principles.

Based on the above discussions, we propose the following hypothesis:

To enable empirical testing, we refine Hypothesis 1 into the following testable sub-hypothesis (Hypothesis 1a):

2.3 Temporality of Logics in Corporate Sustainability

Having said that, the institutional logic theory also predicts the possibility of a transient relationship between ESG and firm value. In other words, market and community logics continuously interact in shaping corporate sustainability. At certain times and in specific contexts, these logics may be in competition, whereas in other instances, they can achieve harmonious coexistence. Although institutional logic is often rigid, it also possesses elements of temporality and malleability, allowing them to be dynamic across different contexts of time and space (Gümüşay et al. 2024). The literature on institutional complexity also suggests that logics are not always in direct conflict; they can coexist or even reinforce each other in specific contexts (Greenwood et al. 2011). This is why some studies have documented a non-monotonic relationship between ESG and firm value (Barnett and Salomon 2012; Darnall et al. 2024).

Nollet et al. (2016) assert that the market reacts negatively in the short term but positively in the long term to firms' ESG involvements. On the other hand, Darnall et al. (2024) argue for the existence of penalty zones, where excessive resource allocation to ESG can turn a positive impact into a negative one. Short-term investors (transactional shareholders) often experience more asymmetric information, leading to noise in their reactions to firms' ESG strategies (Oehmichen et al. 2021). In contrast, long-term investors (durable shareholders) are more receptive to innovation and strategies, as they possess the ability to thoroughly evaluate these factors for sustaining firms' competitiveness, thereby reducing asymmetric information (Connelly et al. 2010).

The inter-temporal shift in the nexus between ESG and firm performance can be attributed to initial investor concerns about greenwashing or doubts regarding the authenticity of ESG commitments (Treepongkaruna et al. 2024). Investors may initially view ESG announcements skeptically, questioning the firm's genuine commitment or the long-term impact of such practices (Du 2015). These ESG efforts may be seen as mere marketing tactics (greenwashing) until firms substantiate them with consistent, tangible results (Zharfpeykan 2021). The initial skepticism toward ESG initiatives may lessen as the positive impact of ESG on firm performance becomes clearer (Barnett and Salomon 2012; Nollet et al. 2016).

The challenge of assessing short-term ESG outcomes contributes to investor caution, leading them to wait for clearer, long-term evidence before positively adjusting stock prices. This hesitancy is particularly pronounced in developing economies, where investors often anticipate adverse effects stemming from weak governmental institutions, lower shareholder protection, insufficient sustainability disclosure and reporting standards, and weak enforcement mechanisms (Julian and Ofori-dankwa 2013). However, CSR literature documents that the practice of greenwashing tends to be relatively lower in developing economies with high market competition as the incentive to gain legitimacy from such activities is relatively low (Lim and Tsutsui 2012; Roulet and Touboul 2015).

Venkataraman et al. (2016) highlight that, though it is challenging to reconcile market and community logic within the context of tribal development initiatives led by NGOs in India, such a convergence is possible when both the beneficiaries and their families gain sufficient trust in the NGO's market-based activities. In the context of ESG activism, investors' concerns about potential greenwashing can similarly be alleviated as they observe clearer links between ESG initiatives and positive financial performance. When concerns about greenwashing are mitigated and ESG initiatives demonstrate tangible benefits, investors are more likely to reassess and incorporate the positive impact of ESG into valuations, effectively ending the “penalty zones.” Here, the market logic can coexist with the community one. Accordingly, we propose the following hypothesis:

Again, for Hypothesis 2 to be empirically tested, we reframe it into the following specific and testable sub-hypothesis (Hypothesis 2a):

3.1 Data

To prove the above hypotheses, this study focuses on developing economies. After filtering for data availability across countries, our dataset consists of 11,757 firm-year observations of non-financial companies listed in 19 emerging countries across Africa, Asia, Europe, and South America, specifically in (1) Argentina, (2) Brazil, (3) Chile, (4) China, (5) Colombia, (6) Egypt, (7) Hungary, (8) India, (9) Indonesia, (10) Malaysia, (11) Mexico, (12) the Philippines, (13) Poland, (14) Russia, (15) Saudi Arabia, (16) South Africa, (17) Thailand, (18) Türkiye, and (19) the United Arab Emirates (UAE). The study utilizes a 10-year historical dataset spanning from 2013 to 2022, the most recent period for which ESG score data provided by the London Stock Exchange Group (LSEG) is largely complete for our sample. The use of ESG scores from LSEG aligns with most ESG studies, such as Albuquerque et al. (2019) and Drempetic et al. (2020). While acknowledging limitations in the methodology for representing corporate ESG performance, this database stands out for providing the most comprehensive ESG data compared to other rating providers (Baran et al. 2022). LSEG ESG scores also demonstrate a high correlation with other leading ESG ratings, such as Kinder, Lydenberg, and Domini (KLD), Moody's ESG (Vigeo-Eiris), MSCI, S&P Global (RobecoSAM), and Sustainalytics (Berg et al. 2022). Table 2 defines our variables and outlines the data sources.

3.2 Methods

We then evaluate the dynamic interactions between market and community logics, that is, Hypothesis 2, by examining the temporality of the above relationship. In so doing, we incorporate the squared term of ESG scores (and its dimensions) into Equations (1–3) to test for non-monotonicity in the relationship between ESG performance and firm value. To ensure the robustness of the nonlinear relationship between the variables, we also employ Lind and Mehlum's (2010) methodology, consistent with previous studies, such as Darnall et al. (2024) and Li et al. (2024). Unlike traditional tests for nonlinearity that focus solely on the sign and significance of the parameters for ESG ( $\widehat{\beta }$) and ESG_SQ ( $\widehat{\delta }$), Lind and Mehlum's (2010) approach evaluates nonlinearity based on three criteria: (i) the turning point ( $-\widehat{\beta }/\left(2\widehat{\delta }\right)$) must lie within the range of observed data $\left[X_{l,}{X}_h\right]$; (ii) the overall Sasabuchi (1980) must be statistically significant; and (iii) the lower bound slope ( $\widehat{\beta }+2\widehat{\delta }{X}_l$) must be negative [positive], whereas the upper bound slope ( $\widehat{\beta }+2\widehat{\delta }{X}_h$) must be positive [negative] for the U-shaped [inverted U-shaped] relationship.

This study also performs a series of robustness checks by (i) excluding countries with a small number of firms, (ii) adjusting for the impact of the COVID-19 pandemic in 2020 and 2021, (iii) replacing TBQ with PBV, (iv) substituting ROE with ROA, and (v) addressing endogeneity issue with the instrumental variable (IV) regression of two-stage least squares (2SLS), consistent with the previous literature (see El Ghoul et al. 2011; Benlemlih and Bitar 2018). The endogeneity issue, in particular, may arise from omitted variable bias, measurement error, or reverse causality, all of which can lead to biased and inconsistent parameter estimates. In the context of the ESG–firm performance nexus, reverse causality is particularly relevant. Although ESG scores can influence firm performance, firm performance can, in turn, affect a company's ESG initiatives and scores (El Ghoul et al. 2011; Benlemlih and Bitar 2018). IV-2SLS is widely regarded as an effective method to address such endogeneity concerns (Wooldridge 2010). Our use of IV-2SLS aligns with established practices in the literature, including studies by El Ghoul et al. (2011) and Benlemlih and Bitar (2018), which also employ this approach to resolve endogeneity issues.

4.1 Descriptive Analysis

Table 3 presents the descriptive statistics of the variables used in the study. On average, firms exhibit moderate ESG performance, with scores of 36.76 for environmental (E), 40.57 for social (S), and 49.24 for governance (G), yielding an aggregate score of 41.64. The G pillar (consisting of CSR strategy, management, and shareholders aspects) has been a long-standing concern for companies, predating the E and S. This precedence is reflected in its comparatively higher score and lower deviation. The E demonstrates the lowest scores, with some firms even obtaining zero. This underscores the urgent need for significant improvement in their focus on emission reduction, innovation, and efficiency aspects. A notable divergence in environmental practices across firms and industries is also evident from its high standard deviation.

The country-level analysis in Table 4 consistently supports these arguments. India boasts the highest average ESG score (62.19) as well as scores in all three dimensions, followed by Colombia (55) and Türkiye (52.52). Conversely, Egypt, Saudi Arabia, and China rank among the lowest ESG performers, with respective scores of 22.68, 30.57, and 34.31. Saudi Arabia records the lowest score in E (21.77), whereas Egypt demonstrates the weakest performance in S (16.90).

The likelihood of multicollinearity in the model is also low, as the independent variables do not exhibit significant correlations with one another (i.e., Person's correlation coefficient $r\ge \mid 0.8\mid$) (see Table 5). Some correlation among ESG dimensions is anticipated due to their interconnected nature. For instance, the correlation between E and S is particularly high at 0.77. In contrast, G diverges more distinctly, showing lower correlations of 0.34 with E and 0.36 with S.

4.2 Dominant Market Logic

Table 6 provides the baseline regression results to test Hypothesis 1. The results of the first step estimation suggest a strong and significant negative relationship between ESG and TBQ at a 1% significance level (Panel 1). The magnitude of this nexus is economically meaningful. To put it into perspective, if the average ESG score increased by 10 units (a 24% increase) from 41.64 to 51.64, the TBQ would decrease from 1.84 to 1.78. In other words, if Saudi Arabia's ESG level were comparable to India's (increasing from 30.57 to 62.19), the TBQ would decrease from 1.83 to 1.64. Assuming a firm has total assets of $100 million and debt comprising 33.33% of its assets, the decrease in TBQ implies a 16.70% reduction in the firm's market capitalization.

This indicates that investors harbor a negative sentiment toward the ESG-related activities conducted by firms, as evidenced by their lowered expectations for firms' future cash flows. This aligns with shareholder theory and legitimacy theory but contrasts with RBV and stakeholder theories. According to shareholder theory, investors believe that an increased focus on ESG could potentially harm the fundamental performance of the firms, diverting them from choices that are most beneficial for shareholders' welfare. From a legitimacy theory perspective, investors perceive this as a negative signal, indicating underperformance in the company's financial bottom line.

These findings also hold for the individual dimensions of E, S, and G, as shown in Panels 2–4, respectively. However, some variations in the marginal effects of each dimension and their level of statistical significance are also observed. All three dimensions negatively impact firm value significantly at a 1% significance level, except for G. In terms of magnitude, the S dimension has the highest impact, followed by the E, whereas G is the least impactful element in the ESG framework. This discrepancy may arise because the G aspect is relatively more well established than the E and S, resulting in limited variations in practices across the firms in the sample. Investors appear to be less concerned about this aspect of ESG, as they perceive firms with reported ESG data, on average, as also performing well in G.

In the second stage of the estimation, we regress the firm's fundamental performance indicator, ROE, on ESG (Panels 5–8). Unlike TBQ, the results suggest a significant positive relationship between ROE and ESG at a 5% significance level. The magnitude of this relationship is also noteworthy. For instance, if the average ESG score increased by 10 units (a 24% increase) from 41.64 to 51.64, the average ROE would also have increased from 12.4% to 13.4%, representing a meaningful 1% improvement in the firm's profitability. To provide a different perspective, assuming Saudi Arabia's ESG level had been comparable to India's (increasing from 30.57 to 62.19), the aggregate profitability of Saudi Arabia's firms would have increased from 11.9% to 15.1%.

Unlike the first-stage results, this finding is inconsistent with shareholder and legitimacy theories but aligns with the RBV and stakeholder theory. It illustrates how adopting a more sustainable approach is beneficial for a company's fundamentals, as it becomes an important additional internal resource that can be utilized to sustain its competitive advantage in the market. While adhering to ESG principles may reduce the traditional investment universe, it opens new opportunities for managers to expand the company into new territories.

The above finding also holds for the three ESG dimensions. The impact of E, S, and G on ROE is positive, although only the impact of the E is statistically significant. This is consistent with our argument that the company's activities related to the ESG are at least not counterproductive to the firm's fundamental performance.

The diverging results of Equations (1) and (2) could not consistently be explained by the four main theories widely used in the literature above. The firm's market capitalization is determined by all available information, including its profitability. If ESG is positively associated with ROE, it should likewise have a positive impact on TBQ, and vice versa. However, our findings suggest the opposite.

To confirm this, we examine the mediating role of ROE on the ESG–firm value nexus. Panel 9 indicates that the adverse effect of ESG on market value remains unchanged even after incorporating ROE. Despite observing a strong relationship between ROE and TBQ, the impact of ESG on TBQ remains significant at a 1% level, with the same magnitude as observed in Panel 1. Consequently, ROE does not serve as a mediator in the relationship between ESG and TBQ, supporting our Hypothesis 1. The result also holds for the three separate ESG dimensions. They consistently mirror the findings of Panels 2–4, with ROE no longer significantly impacting the firm value.

This is where the institutional theory can contribute to offering a more consistent narrative. We attribute these results to the dominance of market logic in investors' perception of firms' ESG initiatives. The dominant market logic of investors hinders their recognition of the beneficial impact of embracing ESG activities within the company. Their expectations are adversely influenced more by the hypothetical possibility of a reduction in investment opportunities associated with adhering to ESG principles, rather than by the actual impact of ESG on the company's fundamental performance.

4.3 Temporality of Competing Institutional Logics

We then examine the temporality dimension of market dominance in corporate ESG observed above. Table 7 suggests a non-monotonic relationship between ESG performance and the firm's market value, consistent with our Hypothesis 2. Both the ESG scores and its squared term are significant at the 1% level. The differing signs of these coefficients indicate a U-shaped relationship, where investors initially penalize the market value of firms engaging in ESG activities. However, as ESG performance improves further, investors begin to appreciate the firm's involvement in ESG activities. The magnitude of the parameters is also economically plausible. Given that the average ESG score is 41.64, negative investor sentiment will persist until the score reaches 57.64, creating transient penalty zones. Beyond this point, investors' attitudes toward ESG become positive. This U-shaped relationship between ESG and firm value highlights the dynamic interactions between market and community logic. Initially, these logics are in competition, with the market serving as the dominant institutional order. However, a state of harmonious coexistence is achieved when $∆\mathit{ESG}\mathit{\text{Performance}}$ reaches 16 (see Figure 1).

The initial reaction of investors is negative as their market logic shapes their perception that many ESG activities are non-income-generating and harmful to the company. Previous research indicates that investors tend to perceive ESG initiatives as greenwashing and a mere marketing practice to cover companies' inability to generate financial profits (Treepongkaruna et al. 2024). They question whether higher ESG performance reflects a genuine commitment toward sustainability that will yield long-term benefits (Du 2015). The difficulty in measuring the immediate effects of ESG initiatives encourages investors to wait for sustained evidence of impact, contributing to a cautious approach to ESG-related valuations. This is why ESG activities may initially be dismissed as mere marketing efforts until companies consistently deliver concrete outcomes (Zharfpeykan 2021). This suggests that investors' initial attitudes toward ESG are influenced by hypothetical logic stemming from market order rather than by its actual financial impacts.

However, as evidence of ESG's positive influence on firm performance becomes more apparent, investor skepticism may gradually decrease (Barnett and Salomon 2012; Nollet et al. 2016). Once companies address greenwashing concerns and prove the value of ESG efforts, investors are more inclined to recognize and integrate these positive effects into firm valuations, ultimately moving beyond the initial “penalty zones.”

To ensure the robustness of the transient penalty zones in the interaction between market and community logics within the ESG performance–firm value nexus, we employ Lind and Mehlum's (2010) methodology on the first- and third-stage regressions presented in Table 8. This method is not applied to the second-stage regression, as no evidence of nonlinearity was identified (see Table 7). Our findings confirm the presence of a U-shaped relationship between ESG performance (E/S/G dimensions) and Tobin's Q. First, the extremum points and Fieller's (1954) confidence intervals for ESG (E/S/G) largely fall within the observed data range. For example, the Fieller confidence interval for turning point Environmental dimension ([4.5, 8.5]) is well above its minimum observed value of zero. Second, the overall Sasabuchi (1980) test is significant at the 1% level in the first- and third-stage regressions, rejecting the null hypothesis of no nonlinear relationship between ESG (E/S/G) and TBQ. Finally, the lower bound slope ( $X_l$) and upper bound slope ( $X_h$) are consistently negative and positive, respectively, across all models in the first- and third-stage regressions, reinforcing the evidence of a U-shaped relationship between the variables. These results collectively support our hypothesis regarding the temporality of institutional logics in ESG practices within the Global South.

4.4 Robustness Checks

Our findings are robust to variations in the (i) sample, (ii) period, (iii) measurements of firm value and fundamental performance, and (iv) endogeneity problem. For the sake of brevity, we present only the robustness test results related to the endogeneity problem, although the other robustness checks are available in Appendix B.

We address potential endogeneity in our models by employing the IV 2-SLS method. We use the mean ESG (and its dimensions) by industry and the initial ESG (and its dimensions) level as IVs in Table 9a (first stage), following the approach of El Ghoul et al. (2011) and Benlemlih and Bitar (2018). These studies provide evidence that these IVs are exogenous to factors influencing the contemporaneous (current) ESG score. The initial ESG score (In_ESG) is considered exogenous because it reflects past ESG performance and is unlikely to be affected by current unobserved factors driving firm performance. The sector-level ESG score (Mean_ESG) captures external ESG trends at the industry level, rather than firm-specific dynamics, making it less likely to be influenced by individual firm performance. Subsequently, we utilize the fitted values of our IVs in Table 9b (second stage) and document consistent results with the baseline estimations.