The Evolution of Environmental, Social, and Governance (ESG) Performance: A Longitudinal Comparative Study on Moderators of Agenda 2030

Abstract

Agenda 2030, embodied by the United Nations’ Sustainable Development Goals (SDGs), represents a global commitment to advancing transparency, accountability, and sustainable development. This study examines whether the introduction of the SDGs in 2015 is associated with changes in environmental, social, and governance (ESG) performance trajectories among major multinational corporations. The analysis uses a piecewise latent trajectory model to examine the ESG trajectories of 320 Global Fortune 500 firms, spanning both the manufacturing and service sectors across developed and developing economies, over the period of 2010–2021. The time frame is deliberately segmented into a pre-SDG period (2010–2015) and one post-SDG implementation (2016–2021) to capture how ESG practices evolved following the launch of the SDGs as a global policy milestone. Our results highlight significant governance improvements in developed economies, especially within manufacturing, driven by regulatory changes and mandatory reporting, while environmental performance trends are more variable and social factors lag in some regions. These findings yield actionable insights for policymakers and managers by pinpointing industrial and regional disparities, thereby informing targeted strategies to advance SDG-aligned ESG practices and harmonize future reporting frameworks.

1. Introduction

The concept of global sustainability was formalized through the United Nations’ 2030 Agenda for Sustainable Development and the Sustainable Development Goals (SDGs), adopted in September 2015 to set concrete targets for ending poverty, protecting the environment, and fostering global prosperity [1]. The Paris Agreement of the same year complemented the SDGs by defining targets for global greenhouse gas reductions and net-zero emissions by 2050 [1]. Together, these frameworks have prompted governments and firms to implement new reporting standards and risk management initiatives, such as the EU’s Carbon Border Adjustment Mechanism and the Equator Principles, to improve transparency and accountability [2,3]. These measures highlight the central role of corporations and investors in achieving sustainability objectives.

While sustainability frameworks are now widely adopted, it remains unclear how the launch of the SDGs in 2015 altered firm-level environmental, social, and governance (ESG) performance, and whether these shifts differ by industry or country context. Corporate accountability for sustainability is increasingly assessed through ESG ratings, which are essential to evaluating value creation and stakeholder engagement, but inconsistencies between rating providers raise questions about their validity [4]. Moreover, ESG ratings’ convergence may obscure underlying differences in sustainability practices.

This study examines how the SDGs and Paris Agreement are associated with ESG trajectories among multinational firms. Using a piecewise latent trajectory model, we analyze ESG performances across 320 Global Fortune 500 firms from 2010 to 2021, intentionally segmenting the analysis into pre-SDG (2010–2015) and post-SDG (2016–2021) periods to capture any policy-driven changes. We specifically test whether ESG trajectory slopes increase following the SDG milestone and whether these trends are moderated by industry type and economic development, in accordance with the mechanisms outlined in Section 2. Firms’ size and age are included as controls to provide a nuanced view of ESG practices across sectors and economies.

Our findings reveal statistically significant improvements in ESG performance—particularly in the governance and environmental dimensions—in developed economies. These results suggest that Agenda 2030’s introduction is associated with observable changes in corporate behavior, underscoring its potential impact on global sustainability practices. However, challenges remain, including methodological inconsistencies in ESG ratings and limited data availability, especially for small and medium enterprises (SMEs). These gaps highlight opportunities for further research into how firms respond to global sustainability regulations across diverse economic contexts.

This study contributes to the ongoing dialog on corporate sustainability by providing empirical evidence of how global policy milestones shape ESG trajectories. The findings offer implications for sustainability theory, corporate strategy, and policymaking, emphasizing the need for collaboration between firms and regulators to achieve shared sustainability goals.

The remainder of this paper is organized as follows: Section 2 reviews the theoretical mechanisms and hypotheses; Section 3 details the data sources and model specification; Section 4 presents empirical findings and slope contrasts; and Section 5 and Section 6 discuss mechanisms, limitations, and policy implications and conclude the study.

2. Theoretical Background and Hypotheses

2.1. The Quest for Sustainable Economic Development

Sustainability has become central in global policy due to concerns about the environmental impacts of economic activities [5,6,7]. System theory provides a valuable lens for analyzing these challenges, emphasizing the interconnectedness of ecological, social, and economic systems [8,9,10]. Recent systematic reviews highlight that agency theory, stakeholder theory, and legitimacy theory are the most prominent conceptual foundations guiding contemporary ESG research. These frameworks help clarify how governance structures, board diversity, and stakeholder engagement interact with ESG outcomes across different sectors and regions and are widely adopted by high-quality studies in non-financial industries [11]. Foundational work in ecological economics highlights how microeconomic actions can collectively exceed the macroeconomy’s carrying capacity, stressing the need for sustainable development to avoid ecological collapse [12,13]. Evolutionary and ecological economics further explore how economic growth must align with environmental limits rather than being used to pursue perpetual expansion [6,14]. The growing consensus is that the Earth’s tolerance for pollution and over-exploitation is nearing its limit [8,9], prompting a shift toward policies that prevent the human ecological footprint from exceeding planetary boundaries [5]. It has been argued that the end of unrestrained economic growth marks the beginning of human transformation [12,15]. Ecological economics, in particular, focuses on the interplay between economic growth, environmental quality, economic activities, carrying capacity, and environmental resilience [16].

2.2. Corporate Social Responsibility (CSR) and the Policy Perspective

Corporate Social Responsibility (CSR) serves as a foundation for corporate sustainability but differs in its ethical roots and normative focus [17]. CSR emphasizes moral obligations and accountability, while corporate sustainability adopts a systems-based, empirically driven approach that is rooted in systems theory and the United Nations Sustainable Development Goals (SDGs), prioritizing dynamic management of economic, social, and environmental interdependencies to ensure long-term viability [18,19]. Although both concepts intersect in addressing the business–society relationship, their paradigms remain distinct: CSR is prescriptive and value-driven, whereas sustainability analyzes interdependencies empirically [17].

The evolution of CSR is marked by foundational theories and frameworks, including corporate social performance [20], stakeholder theory [21], and integrative approaches [22,23]. Over time, the convergence of CSR and sustainability has led to overlaps in constructs and measurement tools, driven by a shift toward a business case logic [17]. While this integration enhances managerial relevance, it risks diluting the unique theoretical and practical insights of each paradigm. A growing body of research further critiques CSR for its persistent ambiguity, lack of clear economic incentives, and dominance of “instrumental logic”—where social and environmental engagement is often justified primarily as a means to reputational or financial gain—potentially undermining substantive sustainable development [22,23,24,25].

Corporate citizenship, which is closely related to CSR, has also faced criticism for its limited theoretical foundation and use as a tool for enhancing prestige rather than addressing systemic risks [18,19,25]. Matten and Crane [26] present critiques of corporate citizenship, while Banerjee [27] discusses the concept’s limits and instrumental logic. The “insurance-like” effect of CSR activities is recognized [28,29], but their potential to drive sustainable development and effective stakeholder alignment remains underexplored [24,30].

Thus, while CSR and sustainability frequently overlap in managerial discourse, their theoretical, ethical, and systemic roots require clear differentiation to avoid persistent conceptual ambiguity and guide effective policy, research, and management practice [17,18,19,22,23].

2.3. ESG and Its Role in Sustainability

Systematic literature reviews, such as that by Ed-Dafali, Adardour, Derj, Bami, and Hussainey [11], show that agency, stakeholder, and legitimacy theories dominate recent ESG research—explaining governance, board composition, and institutional impacts on ESG outcomes. These reviews also recommend integrating multiple theoretical frameworks to enhance the nuance and policy relevance of ESG analyses, especially in varied industrial and regional contexts. The introduction of global policy frameworks in 2015 triggered different adjustment mechanisms for each ESG pillar. Environmental performance may have improved post-2015, especially where green capital investment, sector-specific regulation, and global environmental measures create incentives for action—effects that are primarily expected in capital-intensive manufacturing and advanced economies (e.g., CBAM, EU Green Deal). Social performance is anticipated to show weak or inconsistent post-2015 slope changes, as the standards remain diffuse, outcomes heterogeneous, and regulatory convergence limited (see systematic reviews on “S” measurement divergence). Governance improvements are expected to be the most immediate effect following the SDGs due to convergence on board monitoring, stricter disclosure, mandatory reporting, and new stewardship codes—particularly in advanced economies and the manufacturing sector (see, e.g., the European Union’s CSRD and stewardship code research).

ESG initiatives have emerged to bridge CSR and environmental sustainability, especially in risk management and alignment with the SDGs [21,31,32]. Although ESG and CSR have evolved without standardized metrics, they provide a pathway for firms to align with international policy frameworks and measure progress toward the SDGs [33,34]. The rise of ESG reflects a growing interest in the “business case” for sustainability, moving beyond philanthropy to focus on value creation and accountability [24,25,35].

Rating agencies have developed various scoring systems for ESG practices, but the connection between these scores and the SDGs remains inconclusive, complicating firms’ efforts to track their contributions to global goals [36,37]. Establishing a clear link between ESG and SDG scores is essential for firms to evaluate their impact, enhance accountability, and report progress to stakeholders [34,36,38]. Scholars have begun mapping SDG targets to ESG variables at the firm level, emphasizing the importance of integrating environmental, social, and economic considerations for sustainable development [39]. As the limits of traditional single materiality become clear, the concept of “double materiality”—evaluating both the firm’s impacts on the world and the world’s impacts on the firm—is emerging as a critical principle for robust, credible ESG-SDG integration and reporting [40]. Despite these regulatory advances, significant gaps remain in the standardization and consistent evaluation of “S” and “ESG” performance, especially in linking firm activity to broader social outcomes and just transition goals [40].

Firms aiming to advance sustainable development must integrate environmental, social, and economic considerations, aligning their ESG practices with the SDGs. Prioritizing sustainable economic growth alongside stakeholder interests enables organizations to identify and measure the interconnections between ESG and the SDGs, supporting multi-dimensional value creation. This integrated approach offers a practical pathway for companies to assess their progress toward global sustainability objectives. Based on this rationale, we posit the following:

Hypothesis 1.

For each ESG pillar, the rate of improvement (trajectory slope) is greater after 2015 than before, especially for governance.

This hypothesis aligns with major empirical findings that have been synthesized in recent ESG literature reviews, which document time-based shifts in ESG performance following major policy events such as the SDG launch, especially when analyzed through the lens of stakeholder accountability and governance structures [11].

2.4. The Impact of ESG on Corporate Performance

The diversification of ESG reporting metrics and lack of standardization present significant challenges for comparing ESG performances across firms [37,41]. Recent regulatory initiatives reflect global efforts to harmonize ESG reporting, with the European Union’s Corporate Sustainability Reporting Directive (CSRD) and the International Sustainability Standards Board (ISSB) representing two major approaches. The CSRD mandates comprehensive sustainability reporting based on “double materiality”—requiring companies to disclose both how sustainability issues impact financial performance and how their activities impact society and the environment—while enforcing external audits and expanding coverage to a wide range of entities [42]. In contrast, the ISSB focuses mainly on financial materiality for investors, offering a streamlined “global baseline” but with less stakeholder engagement [42]. Despite these advances, both academic and practitioner communities highlight persistent challenges: the proliferation of frameworks, lack of standardized indicators, and methodological inconsistencies continue to impede comparability, reliability, and cross-country benchmarking of ESG outcomes [43,44]. Concerns are also increasing that superficial compliance (“greenwashing”) may undermine the real-world impact of ESG disclosure, unless robust global frameworks and auditing mechanisms are implemented [44].

Recent management scholarship stresses the urgent need for further harmonization, stronger assurance practices, and improved data auditability—especially as stakeholders rely on ESG data for diverse decision-making processes [43]. Despite these issues, ESG reporting has gained prominence as stakeholders demand greater transparency in non-financial corporate activities [41,45]. Various frameworks and indices offer benchmarking tools, but differences in criteria and methodologies hinder their comparability and reliability [37,41].

Recent research identifies three main sources of divergence in ESG ratings, scope, measurement, and weighting, with measurement divergence being the most significant [46]. These inconsistencies, including the “rater effect,” undermine the use of ESG ratings for investment decisions and empirical research. Calls for greater transparency and standardization in ESG methodologies are growing, alongside recommendations for improved data collection and collaboration between firms and rating agencies [47].

The CSRHub database, selected as the data source for this study, aggregates ESG ratings from a wide range of providers, offering a comprehensive and neutral perspective on corporate sustainability. Its methodology synthesizes data across industries and countries, addressing some challenges of rating divergence and providing reliable consensus scores for ESG performance.

Governance scores are theoretically expected to improve more rapidly post-SDGs because governance metrics are standardized, subject to regulatory scrutiny, and frequently targeted by reforms such as mandatory disclosure and anti-corruption measures [48,49]. According to recent reviews, organizations respond swiftly to external pressures affecting governance, in line with institutional theory and empirical observations of rapid compliance in developed economies [48]. In contrast, environmental and social dimensions often lag due to operational complexity, voluntary standards, resource intensity, and varied regulatory enforcement, which is consistent with comparable empirical findings [50,51].

2.5. Explaining Social Pillar Challenges via Stakeholder and Institutional Theories

The social pillar of ESG remains the most variable and challenging to regulate, a phenomenon that is best explained by stakeholder theory and institutional theory [48,52,53]. Stakeholder theory recognizes the plurality and diversity of interests—ranging from employees and local communities to customers—whose heterogeneous and sometimes competing claims complicate the establishment of standardized social metrics and practices [48]. Institutional theory further argues that the quality of local institutions, such as the rule of law, education, and economic freedom, determines the ability of organizations to implement and disclose robust social responsibility practices [53]. Where institutional frameworks are strong and stakeholder activism is prominent, as in developed service sectors, companies are incentivized and/or pressured to advance social responsibility, leading to measurable improvements in social responsibility outcomes [52]. In contrast, developing regions often lack the necessary institutional capacity and stakeholder mechanisms for effective social responsibility, contributing to variability and decline in social pillar scores [48,53].

Empirical studies confirm that socially specific drivers—including regulatory enforcement, stakeholder engagement, and institutional quality—account for disparities in the service sector’s social responsibility [54,55]. In developed economies, high institutional quality and active stakeholder participation foster improvement in social outcomes [54]. In comparison, developing economies often face persistent obstacles such as weak regulatory regimes, resource scarcity, and limited stakeholder voice, resulting in declining social pillar performance [55,56].

2.6. Bridging the Gap: Empirical Evidence on ESG, SDGs, and Corporate Performance

Various scholars highlight the strategic role of corporations in advancing sustainable development. Freeman [57] posits that firms should create value for all stakeholders, not solely shareholders, integrating social responsibility with core business strategy as a central tenet of modern governance. Building on this perspective, Chandler [58] emphasizes that embedding CSR into corporate strategy enables sustainable value creation. ESG practices operationalize this alignment. For instance, disclosure of Scope 1–3 carbon emissions aligns with SDG 13 (Climate Action), while stronger governance transparency supports SDG 16 (Peace, Justice, and Strong Institutions).

The Global Reporting Initiative [59] provides a comprehensive framework that facilitates the integration of ESG practices with the SDGs, enabling firms to systematically map and disclose their contributions. Extending this view, Delgado-Ceballos, Ortiz-De-Mandojana, Antolín-López, and Montiel [39] propose a framework that underscores the concept of “double materiality,” noting that although the SDGs are defined at the societal or national levels, corporations play a pivotal role in their realization.

Building on ESG disclosure, impact accounting (IA) advances corporate sustainability reporting by quantifying and monetizing social and environmental impacts. This approach requires firms to identify which of the 17 SDGs and 169 associated targets are most relevant to their industry and value chain, link them to specific indicators, and develop measurable metrics. These impacts can then be converted into monetary values for integration with financial statements, as exemplified by the Impact-Weighted Accounts Initiative [60]. Moreover, the Financial Accounting Standards Board (FASB) explores how climate-related risks can be translated into comparable and reliable financial reporting under existing accounting standards. For example, accounting treatments for carbon credits may require firms to assess whether extreme weather events trigger asset impairments or whether potential fines or carbon taxes from unmet emission targets should be recognized as liabilities [61]. Collectively, these frameworks demonstrate how ESG practices, when combined with impact accounting, provide a practical and measurable pathway through which corporations can operationalize the SDGs and contribute tangibly to global sustainability.

Extensive research has examined the link between ESG performance and corporate financial performance (CFP), with evidence from both developed and developing economies indicating a generally positive relationship [62,63,64,65,66,67,68,69,70,71]. Profitable, larger, and less leveraged firms tend to excel on sustainability measures and align their ESG activities with the SDGs [38,65]. ESG disclosure is associated with reduced debt financing costs and improved access to financial resources, supporting green financing and environmental protection objectives [72,73]. ESG has become a crucial form of non-financial disclosure for firms, enhancing transparency and efficiency, particularly in governance [65,69,72,74]. It is also recognized as a key risk factor, with strong ESG performance being linked to lower value at risk and higher returns [35,64,66,75,76,77]. However, the recent literature increasingly spotlights the critical role of regulatory mandates and reporting standards—such as the EU Non-Financial Reporting Directive (2014/95/EU), the forthcoming Corporate Sustainability Reporting Directive, and national disclosure laws—in accelerating post-2015 advances in ESG transparency and corporate accountability. For example, Papa et al. [78] provide robust empirical evidence that the EU Directive led to a significant rise in both the quantity and comparability of environmental disclosures among Italian firms, with the greatest impacts being on information that is explicitly required by law and increased adherence to GRI standards. Their results underscore that well-designed non-financial disclosure laws can substantially improve the quality of ESG reporting and the alignment of firm practices with regulatory expectations, demonstrating the importance of such mandates, especially in advanced economies.

The benefits of ESG investments are realized in the long term, and empirical studies suggest that consistent ESG expenditure improves sustainability performance and resilience during crises [62,63,79,80,81,82,83,84].

Industry type and economic development are important moderators in the ESG–CFP relationship, with research highlighting the need for policy-relevant insights to inform strategies in the context of the 2030 SDGs [70,85,86]. Industry moderates ESG slope changes, as manufacturing is more affected by direct regulatory mandates, environmental process requirements, and internal controls, while service sectors often externalize impacts and face more diffuse or voluntary standards. Economic development moderates the translation of SDG frameworks: advanced economies possess greater institutional capacity, regulatory enforcement, and investor protection, resulting in a more pronounced and efficient shift in ESG slopes after 2015. Emerging markets, where regulations are voluntary or fragmented, often see weaker or delayed changes. Given the persistent regional and sectoral differences in ESG drivers and barriers, recognizing these moderating effects is vital for tailoring sustainability strategies to the 2030 SDG context [43,46]. As emphasized by Ed-Dafali, Adardour, Derj, Bami, and Hussainey [11], robust governance, board diversity, and strategic integration of ESG principles are critical drivers of improved ESG performance and profitability. Their comprehensive framework for non-financial firms demonstrates how the industry context moderates these effects, supporting the rationale for segmenting our analysis by industry type and economic development in this study. This study simplifies industry classification and focuses on the moderating effects of industry and economic development on ESG performance trajectories, aiming to provide actionable guidance for corporate sustainability, in alignment with global goals. Figure 1 illustrates our research model, capturing these policy-driven dynamics, and visually represents these moderating effects on the trajectory of ESG performance across time, industry, and economy.

By applying key conceptual frameworks, highlighted by Ed-Dafali, Adardour, Derj, Bami, and Hussainey [11], this study extends prior systematic literature review findings by empirically testing the moderating effects of industry and economic context, thereby contributing nuanced evidence to ongoing debates on ESG performance measurement, governance, and SDG alignment. Rather than focusing solely on single-market or economy-level analyses, we adopt a piecewise comparison based on economic development levels, utilizing panel data. This approach enables us to examine how both industry type and economic development moderate the trajectory of ESG performance over time, particularly in the context of the 2030 SDGs. Based on this framework, we propose the following hypotheses:

Hypothesis 2.

Industry moderates slope changes, with greater post-2015 slope increases in manufacturing sectors than in services.

Hypothesis 3.

Economic development moderates slope changes, with advanced economies exhibiting larger post-2015 increases than developing economies.

By examining these hypotheses, our research aims to provide policy-relevant insights into the evolving landscape of corporate sustainability in the context of global sustainability goals.

3. Data and Methods

3.1. Data

We obtained ESG data from the CSRHub database, a credible source aggregating information from 350 million sustainability- and CSR-related sources [4,87]. Financial data were sourced from Compustat. To mitigate ESG rating divergence, CSRHub employs a consensus aggregation methodology, harmonizing data from more than 900 expert sources. This process normalizes disparate indicators according to GRI-based categories and applies systematic weighting to ensure cross-source comparability. Empirical studies confirm that aggregation significantly reduces rater-specific biases and enhances the reliability of composite ESG scores. For example, Conway [88] demonstrates statistically robust agreement between CSRHub’s consensus ratings and Bloomberg data across social, environmental, and overall categories, especially for large and medium-sized firms. Berg, Kölbel, and Rigobon [46] show that aggregation and common taxonomies directly address scope, measurement, and weighting divergence in ESG ratings, supporting the methodological reliability of consensus-based approaches in large-sample studies.

To ensure coverage, comparability, and methodological rigor in our ESG measurement, we selected the CSRHub database as our principal data source. The selection of CSRHub is further validated by multiple recent peer-reviewed empirical studies that demonstrate its reliability, research acceptance, and diverse applicability. Empirical research documents the robustness of CSRHub’s ESG ratings in cross-sectoral and international analyses. Thiart [89] applies CSRHub data to investigate the relationship between ESG rating and tax transparency across 112 South African listed firms, finding the aggregated ratings both credible and suitable for regulatory research contexts. Vaughan [90] utilizes CSRHub’s human rights and supply chain scores in a five-year benchmarking analysis of U.S. hospitality and tourism companies, confirming the operational credibility and industry relevance of CSRHub ratings for advanced social responsibility assessments. Sandberg, Alnoor, and Tiberius [4] extend CSRHub’s empirical reach, employing its ESG scores to analyze the financial impact of sustainability performance in the European food industry and corroborating the usefulness of the database as an academically accepted data source for sectoral and financial comparisons. These examples illustrate CSRHub’s established role in advanced sustainability and social responsibility research. The database’s systematic harmonization of multi-source data and demonstrated validity in published studies support its appropriateness for large-scale, cross-industry ESG analyses.

The study sample comprises 320 large-cap firms, drawn from the Global Fortune 500 list and representing a diverse range of industries and countries. This selection enables robust cross-sector and cross-country comparison, as Fortune 500 companies systematically report using established ESG frameworks and disclosure standards. Fortune 500 companies are selected due to their systematic ESG reporting, established disclosure standards, and the availability of consistent, comparable ESG data on the global scale. While this enables robust analysis and benchmarking, small and medium enterprises (SMEs) are excluded, because standardized ESG disclosures and comprehensive data for SMEs remain limited. Thus, our findings apply specifically to large corporations facing rigorous investor and regulatory scrutiny.

The chosen observation period is 2010–2021, strategically set to exclude years that were impacted by the global financial crisis (2008–2009) and COVID-19 pandemic (2020–2022), thus minimizing bias from exceptional macroeconomic shocks. The time segmentation—with 2010–2015 as the baseline and 2016–2021 as the post-SDG-implementation window—directly corresponds to the launch of the United Nations’ 2030 Agenda for Sustainable Development in 2015, which instituted the SDGs as a global benchmark for corporate sustainability. This design enables a direct empirical evaluation of how ESG practices among the world’s largest multinational firms responded to Agenda 2030’s introduction and subsequent regulation trends. This study’s methodological approach—focusing on Fortune 500 firms and using a panel design and piecewise latent trajectory modeling—reflects best-practice recommendations identified by systematic reviews such as that by Ed-Dafali, Adardour, Derj, Bami, and Hussainey [11], which emphasize sectoral diversity, robust governance metrics, and temporal segmentation for analyzing ESG performance trends across non-financial industries. A comprehensive appendix (see Supplementary Materials) provides firm-level details on industry type, economy classification, country, primary industry, and associated SIC codes. This enables full transparency regarding sample construction and the logic underpinning each classification. Figure 2 provides a clear overview of the sample’s composition by both the type of economy and industry, contextualizing the empirical foundation for this study. The total number of firms is 320.

3.2. Methods

3.2.1. The Piecewise Latent Trajectory Model

We used a piecewise latent trajectory model to analyze firm-level ESG performances, capturing dynamics before and after key policy shifts such as Agenda 2030, and enabling an exploration of the impacts of policy on sustainability.

3.2.2. Control Variables

Firm size (total assets, market value), age (years since founding), revenue, and return on equity (ROE) were included as control variables. Market value was calculated by multiplying the current stock price and total outstanding shares. These controls account for organizational scale, maturity, and financial performance, supporting robust analysis of ESG determinants.

3.2.3. Analytical Methods

We employ the piecewise latent trajectory model to capture ESG performance dynamics before and after a transition point, enabling an analysis of policy impacts and firm-level variations [91,92,93]. Although referred to in some research as a “latent trajectory model”, following Ding, Wu, and Chang [91], our specification is a piecewise linear mixed-effects model (growth model), estimated via PROC MIXED, within the hierarchical linear modeling (HLM) framework. “Latent trajectory” refers here to the modeling of unobserved growth curves at the group level and random effects across firms, as implemented in state-of-the-art HLM software. This approach provides insights into time variation, industry nuances, and economic development effects on ESG, supporting policy analysis and strategic planning that are aligned with the 2030 SDGs. The piecewise latent trajectory model is estimated as follows:

$$\begin{array}{c}{\chi }_{itk}={\alpha }_{ik}+{\lambda }_{1t}{\beta }_{1ik}+{\lambda }_{2t}{\beta }_{2ik}+{\epsilon }_{itk}, \hfill \\ Y_{it}=E{R}_{it}, S{R}_{it}, G{R}_{it}, \hfill \\ i=1, 2, ..., 320; t=1, 2, ..., 12; k=1, 2, 3,\hfill \end{array}$$

(1)

where ${\chi }_{itk}$ is the value of ${\chi }_k$, which is the trajectory variable of the kth rating indicator for firm i at time t; ${\alpha }_{ik}$ is the random intercept; ${\beta }_{1ik}$ and ${\beta }_{2ik}$ are the random slopes before and after the United Nation’s Agenda 2030 SDGs set the trajectory for firm i, associated with the kth rating indicator, respectively; ${\lambda }_{1t}$ and ${\lambda }_{2t}$ are fixed loading coefficients for measuring the time period before and after goal setting; and ${\epsilon }_{itk}$ is the corresponding error. We defined ${\lambda }_{1t}$ and ${\lambda }_{2t}$ with a dummy variable for time = 0 as T1 for a period before the SDGs were implemented between 2010 and 2015, and time = 1 as T2 after the implementation of the SDGs between 2016 and 2021. To capture segment-specific slopes (piecewise linear trajectories), we followed the approach of Ding, Wu, and Chang [91] and coded the time variables so that for the pre-SDG period (2010–2015), T1 represents the number of years from the segment start (coded as −5, −4, −3, −2, −1, and 0), while T2 is set to zero for this period. For the post-SDG period (2016–2021), T2 is incremented by year (coded as 1 to 6), while T1 is set to zero. This construction of T1 and T2 as segment-specific time loading vectors allows us to estimate distinct linear slopes before and after the implementation of the SDGs. By doing so, the model identifies changes in the rate of ESG performance (slope shifts) at the 2015 policy breakpoint, rather than only modeling average level changes. This specification follows best practices in piecewise growth modeling and ensures that we can directly compare pre- and post-policy trajectory slopes within a mixed-effects framework. All estimated contrasts and moderator effects correspond to the proposed hypotheses defined in Section 2. All reported planned contrasts and moderator effects were pre-specified and correspond to the proposed hypotheses. No post hoc adjustments were applied.

Among the four ESG indicators defined in the CSRHub schema—Community, Employees, Environment, and Governance—this study operationalizes ESG performance using three top-level ratings: Environmental Rating (ER), Community Rating (SR), and Governance Rating (GR). Specifically, the Environment category is used as the ER to represent the “E” dimension, capturing a firm’s interactions with the environment at large, including its use of natural resources and impact on ecosystems. The category evaluates corporate environmental performance, compliance with environmental regulations, mitigation of environmental footprint, leadership in addressing climate change through relevant policies and strategies, energy-efficient operations, and the development of renewable and alternative technologies [87]. It also covers disclosure of sources of environmental risks and liability, implementation of conservation and pollution prevention programs, demonstration of sustainable development strategies, integration of environmental responsiveness with management and the board, and programs to engage stakeholders for environmental improvement. The Community category is mapped to the SR, reflecting a firm’s impacts on community development, product responsibility, human rights, and supply chain management. Finally, the Governance category is adopted directly as the GR, which covers aspects of board structure, leadership ethics, transparency, and stakeholder relations. While the indicator of Employees is available in the CSRHub schema, the three-dimensional ESG operationalization used here aligns with the most relevant and robust CSRHub indicator categories for this analysis.

The initial dataset comprised 6613 firms. After removing firms without ESG scores reported in at least one year during 2010–2021, 1719 unique firms remained. To ensure robust comparability and minimize bias due to missingness, firms or firm-years lacking any of the three ESG pillar scores were first removed, yielding a reduced sample of 369 firms with at least some available year-level scores. Next, we incorporated five control variables (total assets, market value, ROE, firm age, revenue). If a firm lacked the required control variable data for a given year, only that specific firm-year observation was excluded, rather than the firm’s entire record. This further filtering resulted in a final panel of 320 firms with complete indicator and control variable data for all included years. The final sample constitutes an unbalanced panel with firm-year observations distributed across 2010–2021; however, the majority of firms have nearly complete annual coverage. To mitigate the influence of extreme values and potential data errors, all continuous variables (ESG pillar scores and financial controls) were winsorized at the 1st and 99th percentiles by year and pillar. This follows best practice in ESG panel studies to guard against distortion from outlying observations while preserving the majority of the empirical variation.

This study analyzes the piecewise latent trajectory model using three indicators, ER, SR, and GR, each ranging from 0 to 100. ER comprises the environment pillar score, reflecting a company’s impact on natural resources, ecosystems, regulatory compliance, climate policy, energy efficiency, and sustainability initiatives. SR includes the social pillar score, which factors in community–human rights, supply chain, product quality and safety, product sustainability, community development and philanthropy, employee diversity, labor rights, union treatment, compensation, benefits, training, health, and worker safety. GR is defined as leadership ethics, board composition, executive compensation, transparency, reporting, and stakeholder treatment.

To detect the error structure, this study employs the first-order autoregressive AR(1) model (e.g., [94], p. 175). The structural part of the model with random intercept (${\alpha }_{ik}$) and random slopes (${\beta }_{1ik}$ and ${\beta }_{2ik}$) is presented as follows:

$$\begin{array}{l}{\alpha }_{ik}={\gamma }_{00}+{\gamma }_{01}\mathrm{DInd}{1}_i+{\gamma }_{02}\mathrm{DEcon}{2}_i+{\zeta }_{0i},\\ {\beta }_{1ik}={\gamma }_{10}+{\gamma }_{11}\mathrm{DInd}{1}_i+{\gamma }_{12}\mathrm{DEcon}{2}_i+{\zeta }_{1i},\\ {\beta }_{2ik}={\gamma }_{20}+{\gamma }_{21}\mathrm{DInd}{1}_i+{\gamma }_{22}\mathrm{DEcon}{2}_i+{\zeta }_{2i},\end{array}$$

(2)

where all gammas are fixed-growth parameters, and ${\zeta }_{0i}$, ${\zeta }_{1i}$, and ${\zeta }_{2i}$ are level-2 error variances. In the level-2 model, two time invariant predictors are used. Two dummy variables, industry type and economy type, are created and defined as follows: DInd1 = 1 for service and =0 for manufacturing; DEcon2 = 1 for developing countries and =0 for developed countries.

Industry type was coded as 0 for manufacturing and 1 for services, based on each firm’s primary source of value creation and business model. This classification approach follows the precedent in cross-sectoral ESG studies and draws upon recent systematic reviews [11], as well as OECD and World Bank sectoral definitions [95,96]. Firms categorized as manufacturing (DInd1 = 0) are those that are primarily engaged in the physical production, processing, or transformation of goods, including mining, construction materials, chemicals, and equipment. In contrast, firms whose principal activities are in finance, insurance, retail, wholesale, information technology, telecommunications, transportation, or the provision of consumer and specialized business services were coded as services (DInd1 = 1), in accordance with international business statistics guidelines. For conglomerates and firms with diversified operations, the classification was assigned based on the segment contributing the largest share of consolidated group revenue, informed by company reports and widely accepted research practice. This hybrid approach, while not strictly SIC-driven, aligns with operational and revenue-focused distinctions that have been recommended in the recent ESG measurement literature. Economy classification was conducted according to the International Monetary Fund (IMF) World Economic Outlook, April 2025, and the World Bank country and lending group classifications for FY2025–2026, with economies classified as “developed” (advanced/high-income) or “developing” (emerging, upper-/middle-/lower-income) based on published annual lists. Under these definitions, Hong Kong is treated as developed and China as developing, which is consistent with both the IMF and World Bank taxonomies.

To strengthen the robustness of our analysis and account for potential moderating factors in ESG dimensions, we include firm-level characteristics—namely, firm size, market value, and age—as control variables, given their documented influence on ESG performance and sustainability disclosure [32,97]. A firm’s size, market value, and age are standard control variables that have been shown to influence ESG disclosure and performance due to their relationship with organizational legitimacy, reputation, and resource availability [32,98,99]. Firm size, measured by total assets or market capitalization, is associated with higher ESG scores and reporting quality, primarily because larger firms have more resources and external pressure to disclose [32,100]. Larger firms tend to attract more external scrutiny, possess greater legitimacy and reputational leverage, and maintain more resources to support effective ESG data provision and transparency. Similarly, firm age is routinely controlled as a proxy for organizational maturity and historical legitimacy [35,99,100]. This control variable is crucial for understanding how a firm’s longevity and experience in the market may influence its ESG performance. As the policy landscape evolves over time, older firms may possess historical perspectives and practices that affect their sustainability approach and response to policy-driven objectives.

In summary, our comprehensive methodology not only scrutinizes the temporal evolution of ESG performance but also meticulously accounts for moderating factors—such as firm size and age—to ensure a nuanced and policy-relevant analysis of corporate sustainability practices within a dynamic global context. Through this rigorous approach, we aim to provide tangible policy implications to guide policymakers, businesses, and stakeholders in navigating the evolving landscape of ESG performance, aligning with the objectives of sustainability and the United Nations’ 2030 Sustainable Development Goals (SDGs), established in 2015. This alignment, grounded in thorough analysis, forms the foundation of our research’s contribution to policy-oriented discourse.

4. Results

Table 1. Descriptive statistics of firm-level variables (N = 320).

Variable	N	Mean	StdDev	Min	Max
ER	4387	58.98	9.86	14	86
SR	4388	54.46	8.11	18	88
GR	4391	52.9	8.25	23.33	85
Industry type	4428	0.49	0.5	0	1
Economy type	4428	0.18	0.38	0	1
Total assets	4140	267,737.11	561,212.76	0.3	5,108,631.27
Market value	3840	65,549.58	119,848.25	6	2,413,423.40
ROE	3978	14.96	35.44	−240.02	1048.62
Firm age	4428	77.38	51.27	8	357
Revenue	4143	61,769.91	57,618.21	201.43	559,151

Notes: Industry type and economy type are binary-coded (0 = manufacturing/developed countries, 1 = services/developing countries). ER = Environmental Rating (Environment); SR = Social Rating (Community); GR = Governance Rating (Governance).

reports descriptive statistics. Our analysis, employing the piecewise latent trajectory model for ESG indicators, yields pivotal insights that have direct policy implications.

Table 2. Main results of the piecewise latent trajectory model analysis.

	ER		SR		GR
Intercept	60.9516	***	52.4265	***	55.2082	***
T1	0.2335		0.0124		0.1940
T2	0.4334	**	0.2357	**	0.4373	**
Industry type	−3.9828	***	−1.4913		−0.8034
Economy type	−4.7989	***	−3.1565		−0.5259
Total assets	0.0007	***	0.0013	***	0.0033	***
Market value	−0.0027		−0.0023		0.0046
ROE	0.0013		−0.0041		0.0002
Firm age	0.0128		0.0149	**	0.0174	**
Revenue	−0.0100		0.0000	***	0.0000	***
T1 × Industry type	0.3629		−0.4024		0.2041
T2 × Industry type	0.6545	***	0.2537		−0.1481
T1 × Economy type	1.3873	***	1.6683	***	1.5579	***
T2 × Economy type	−0.1397		−0.3565	***	−0.6502	***
T1 × Total assets	0.0005	***	0.0000	***	0.0000	***
T1 × Market value	−0.0012		0.0000		0.0007
T1 × Firm age	0.0013		−0.0027		−0.0006
T1 × Revenue	0.0026		−0.0051		−0.0030
T1 × ROE	−0.001		−0.0026		−0.0012
T2 × Total assets	0.0003	***	−0.0002	***	−0.0004	***
T2 × Market value	0.0008	***	0.0005	***	−0.0009	***
T2 × Firm age	−0.0003		−0.0021	*	−0.0023	*
T2 × Revenue	0.0000		0.0000	***	0.0000	***
T2 × ROE	0.0009		0.0029		0.0001
Observations	3679		3684		3683
Log likelihood	23,480.5		22,736.0		21,245.6
Akaike Information Criterion (AIC)	23,542.5		22,798.0		21,307.6
X2	2561.77	***	2506.89	***	3984.05	***

Note: Total assets, market value, and revenue are expressed in USD millions in both the data and model. Coefficients thus represent the effect of an increase of USD one million in the respective variable, and significance is denoted as * p < 0.05; ** p < 0.01; and *** p < 0.001. ER = Environmental Rating (Environment); SR = Social Rating (Community); GR = Governance Rating (Governance).

summarizes our main results.

Using SAS PROC MIXED [101] in hierarchical linear models (HLMs), we uncover substantial shifts in firm performance trajectories within the realm of ESG following the adoption of Agenda 2030. Specifically, the three ESG pillar scores show positive improvements from 2016 to 2021 (T2). When considering industry and economy type, it is worth noting that the environmental pillar stands out. The environmental pillar exhibits a significant positive impact from industry type (the mean slope of T2 × industry type is 0.6545, with p < 0.001), underlining the pivotal role of industry-specific practices in advancing environmental sustainability. While no significant impacts are detected for industry and economy type in the social and governance pillars, a notable trajectory emerges at the 0.001 level. On the other hand, the economy type is significantly associated with the three pillars before 2015 (in ER, SR, and GR, T1 × economy type are 1.3873, 1.6683, and 1.5579, respectively, at a significant level of p < 0.001) but negatively associated with social and governance scores after 2015 (T2 × economy type are −0.3565 and −0.6502). To further understand the implications, we conducted a detailed comparison of the mean slopes of the trajectories before and after Agenda 2030 between the periods of 2010–2015 (T1 × industry type and T1 × economy type) and 2016–2021 (T2 × industry type and T2 × economy type). These comparisons reveal the distinct effects in the environmental aspect from the social and governance dimensions. These findings underscore the significant changes occurring before and after Agenda 2030 on ESG dimensions and emphasize the need for a comprehensive comparative analysis of piecewise linear trajectories to gain a deeper understanding.

When analyzing the mean slopes of the performance trajectory concerning ESG performance, interesting results emerged. Null hypotheses were tested to examine the mean slopes of the trajectories of the performances of the three scores before and after the implementation of Agenda 2030. As shown in

Table 3. Comparison of mean trajectory slopes for ESG performance before and after the 2030 Agenda.

	Before (2010–2015)	After (2016–2021)	After–Before
ER	0.8149	0.9074	0.0925
SR	0.4857	0.2466	−0.2390
GR	−0.3019	0.2401	0.5419 ***

Note: *** p < 0.001. ER = Environmental Rating (Environment); SR = Social Rating (Community); GR = Governance Rating (Governance).

, the slope trajectory of the governance scoring performance improved after the adoption of Agenda 2030, whereas the changes in the other two pillars were found to be insignificant. Notably, while governance performance exhibits an insignificant mean slope before and after implementation, the difference between these periods (after–before) is positively significant (0.5419, p < 0.001).

A more detailed examination of the mean trajectory slopes for ESG performance, including a matrix before and after the implementation of the 2030 Agenda, is provided in

Table 4. Detailed comparisons by industry type of mean trajectory slopes for ESG performance before and after the 2030 Agenda.

	Before (2010–2015)		After (2016–2021)		After–Before
ER
Developed Economy
Manufacturing Industry	0.5964	*	1.0879	***	0.4915
Service Industry	0.2335		0.4334	*	0.1999
Developing Economy
Manufacturing Industry	1.9836	***	0.9482	***	−1.0354	*
Service Industry	1.6207	***	0.2937		−1.3270	**
SR
Developed Economy
Manufacturing Industry	0.3981		0.2892	*	−0.1089
Service Industry	0.1940		0.4373	**	0.2433
Developing Economy
Manufacturing Industry	1.9559	***	−0.3610	***	−2.3169	***
Service Industry	1.7518	***	−0.2129		−1.9647	***
GR
Developed Economy
Manufacturing Industry	−0.3900		0.4894	***	0.8794	**
Service Industry	0.0124		0.2357		0.2233
Developing Economy
Manufacturing Industry	1.2783	***	0.1329		−1.1454	**
Service Industry	1.6807	***	−0.1208		−1.8015	***

Note: * p < 0.05; ** p < 0.01; *** p < 0.001. ER = Environmental Rating (Environment); SR = Social Rating (Community); GR = Governance Rating (Governance).

, with a focus on results across industry types (D = 0, 1; 0 for manufacturing industry and 1 for service industry) and economy types (E = 0, 1; 0 for developed economy and 1 for developing economy).

Further analysis using the moderated piecewise growth model reveals significant trends in the ER scores across different industry types and economies. Developed economies, particularly in the manufacturing industry, exhibit sustained positive trends in ER performance from 2010 to 2021, with the mean slope increasing from 0.5964 to 1.0879. A similar pattern emerges in the service industry in developed economies, where the mean slope rose from 0.2335 to 0.4334. This indicates a sustained improvement in ER performance among firms in developed economies over the past 12 years.

Conversely, firms in developing economies, both the manufacturing and service industries, exhibit a significant decline in ER after 2015 (after–before = −1.0354, with p < 0.05, and −1.3270, with p < 0.01, respectively). Interestingly, they showed a positive trajectory in ER scores both before and after the adoption of Agenda 2030. Specifically, in the manufacturing industry, ER scores were significantly positive before the Agenda’s adoption (1.9836, p < 0.001) but decreased afterward (0.9482, p < 0.001). Similarly, in the service industry, ER scores were significantly positive before the Agenda’s adoption (1.6207, p < 0.001) but decreased after its implementation (0.2937). This shift marks a challenge in sustainability for developing economies, emphasizing the need for tailored policy responses.

The social rating (SR) score shows mixed results. For manufacturing in developed economies, we recorded a slight improvement through an increase from 0.3981 to 0.2892, indicating a significantly positive trajectory after 2015 (2016–2021) (0.2892, p < 0.05), although the change was not statistically significant. In contrast, the service industry exhibited a significant improvement (0.4373, p < 0.01), with the mean slope increasing from 0.194 to 0.4373, although the growth trajectory (after–before) remained insignificant. The result in the service industry shows that SR performance was positive after 2015, but it still did not exhibit significance in the growth trajectory of the scoring. However, firms in developing economies underperform compared with firms in developed economies across the two periods. In the manufacturing industry, the growth trajectory was significantly positive before the Agenda was introduced (1.9559, p < 0.001), but it became negative after its adoption (−0.3610, p < 0.001), while the service industry shows a similar tendency, with a significant positive trajectory before the Agenda was introduced (1.7518, p < 0.001), becoming negative after its adoption (−0.2129). Both the manufacturing and service industries in developing economies show significantly negative performance trajectories; the tests of the after–before association yielded results of −2.3169 and −1.9647 at the p < 0.001 level. That is, firms in developing economies lag significantly behind in both the manufacturing and service industries, necessitating targeted interventions to bolster their social performance.

Similarly to SR, the GR score of the manufacturing industry in developed economies shows a significant improvement. The mean slope increased from −0.3900 to 0.4894, resulting in a significantly positive performance trajectory, with the after–before calculation yielding a value of 0.8794 at the p < 0.01 level, indicating an improvement in GR after the implementation of Agenda 2030 (0.4894, p < 0.001). The service industry also improved, with the mean slope increasing from 0.0124 to 0.2357, resulting in a positive change of 0.2233, although this change was not statistically significant. In contrast, the GR scoring of the manufacturing industry in developing economies shows a significantly positive growth trajectory before the Agenda was adopted (1.2783, p < 0.001) but turned less positive after its adoption (0.1329), resulting in a significant negative change of −1.1454 at the p < 0.01 level. A similar trajectory was observed in the service industry, with a positive trajectory before the passing of the Agenda (1.6807, p < 0.001), which became negative after its adoption (−0.1208), resulting in a significant negative trajectory and a change of −1.8015 at the p < 0.001 level. These additional results highlight the varying impacts of the 2030 Agenda on ESG performance across different sectors and economies, underscoring the importance of considering the industry and economy type when assessing sustainability initiatives. The fact that manufacturing industries experienced a decline in governance performance calls for regulatory adjustments to enhance governance practices.

When considering the mean slopes of the ESG performance trajectory in terms of industry type and economy type in the $2\times 2$ matrix before and after the adoption of Agenda 2030 in 2015 (Table 3), the overall mean slopes of the trajectory of GR performance exhibit significant improvement after the implementation of Agenda 2030 (with estimates of 0.5419, and p < 0.001). This indicates that Hypothesis 1, proposing an improvement in governance performance, is supported.

Table 4 provides a more detailed analysis. The mean slopes of the GR performance trajectory in the manufacturing sector in developed economies are significantly higher after 2015 than before (with estimates of 0.8794, and p < 0.01), while the slopes in the service sector are statistically insignificant. Furthermore, the mean slopes of the ER performance trajectory in both the manufacturing and service industries in developed economies exhibit significant improvement after 2015 (with estimates of 1.0879 and 0.4334, and p < 0.001 and p < 0.05, respectively). The tests associated with the SR performance of the service sector in developed economies after the implementation of Agenda 2030 are significantly improved, but the figure for the manufacturing industry in developed economies remains unchanged.

Conversely, in developing economies, no clear improvement is observed in the ESG performance trajectory. Rather, it maintains a positive scoring performance in ER, SR, and GR. However, the tests associated with ER for the manufacturing industry in developing economies are positive after 2015 (0.9482, p < 0.001). In general, developing economies exhibit negative mean slopes of trajectory, regardless of the environmental, social, or governance scoring and irrespective of the type of industry. This suggests that developing economies face the challenge of decelerating performance in ESG.

Table 5. Covariance parameter estimates for piecewise growth models.

Parameter	ER	SR	GR
Random effects
Intercept variance (UN)	16.61	26.68	43.57
Slope 1 variance (pre, UN)	1.89	2.18	1.36
Slope 2 variance (post, UN)	0	0	0
Intercept–Slope 1 covariance (UN)	−3.38	−1.32	−2.82
Intercept–Slope 2 covariance (UN)	1.38	−1.24	−3.87
Slope 1–Slope 2 covariance (UN)	0.66	0.17	0.52
Serial correlation
AR(1) residual correlation ($\rho$)	0.47	0.31	0.56
Residual variance	36.27	24.17	23.06

Notes: UN(x, y) indicates elements of the unstructured covariance matrix for random effects (intercept, pre-slope, post-slope). All estimates are from mixed-effects models, fit by maximum likelihood (ML) with AR(1) residuals. See Methods section for parameter definitions. ER = Environmental Rating (Environment); SR = Social Rating (Community); GR = Governance Rating (Governance).

presents the covariance parameter estimates for random intercepts, pre- and post-break slopes, and residual AR(1) structure for the piecewise growth models of ER, SR, and GR.

All financial control variables were evaluated for scaling effects and collinearity risk, as recommended in reviewer feedback. The descriptive diagnostics indicate considerable spread and heterogeneity among the control variables (total assets, market value, ROE, firm age, and revenue), with low to moderate correlations between controls. Specifically, the Pearson correlation coefficients were all below 0.43, substantially below conventional thresholds for collinearity concerns ($\mid \gamma \mid > 0.7$).

Variance inflation factor (VIF) statistics for all predictors in the ER, SR, and GR models consistently ranged between 1.01 and 1.27, demonstrating no significant multicollinearity among the size proxies or other financial covariates. These diagnostics confirm that the coefficient estimates are not compromised by redundant information in the control set. All regression models thus employ appropriately scaled and independent predictors, supporting unbiased inference (see

Table 6. Correlations and variance inflation factors (VIFs) of financial controls.

Variable Pair	Observations	Correlation		ER	SR	GR
Total assets				1.06	1.06	1.06
Market value				1.26	1.26	1.26
ROE				1.03	1.03	1.03
Firm age				1.01	1.01	1.01
Revenue				1.27	1.27	1.27
Total assets–Market value	3840	0.09	***	1.06	1.06	1.06
Total assets–ROE	3978	−0.06	**	1.03	1.03	1.03
Total assets–Firm age	4140	−0.05	**	1.01	1.01	1.01
Total assets–Revenue	4137	0.20	***	1.27	1.27	1.27
Market value–ROE	3709	0.14	***	1.26	1.26	1.26
Market value–Firm age	3840	−0.02		1.01	1.01	1.01
Market value–Revenue	3839	0.43	***	1.27	1.27	1.27
ROE–Firm age	3978	0.00		1.03	1.03	1.03
ROE–Revenue	3977	0.02		1.03	1.03	1.03
Firm age–Revenue	4143	−0.02		1.01	1.01	1.01

Note: ** p < 0.01; *** p < 0.001. ER = Environmental Rating (Environment); SR = Social Rating (Community); GR = Governance Rating (Governance).

).

In accordance with reviewer recommendations, all continuous covariates were z-standardized prior to the regression analysis to ensure comparability and mitigate the effects of differing variable scales. Collinearity diagnostics, including Pearson correlations and VIF statistics, were conducted on both raw and standardized versions of the financial controls. The results were highly consistent, with all VIFs remaining low and no evidence of multicollinearity being detected in either specification. In summary, all recommended control scaling and collinearity diagnostic steps have been rigorously applied. The Pearson correlations among controls were low to moderate, and the VIF statistics were consistently below conventional thresholds for collinearity concerns. These results confirm that the coefficient estimates are not compromised by scaling or collinearity, with full details being presented in Table 6.

Further analysis reveals that Hypothesis 3 is supported, indicating significant performance improvements in developed economies compared with developing ones. In contrast, Hypothesis 2 gains conditional support, given the overall performance that we observed in the developing economy, with only SR in the manufacturing industry in developed economies exhibiting improvement after 2015. While the cross-section effects are mixed, industry-specific analysis (Table 4) reveals that the governance and environmental pillars in manufacturing sectors in developed economies have improved significantly, which is consistent with research emphasizing sectoral and national regulatory effects on ESG adoption [50,51].

In summary, our results confirm an improvement in governance performance (Hypothesis 1), particularly in developed economies’ manufacturing industry, while revealing complexities in the social and environmental dimensions. Developing economies face challenges in all three ESG dimensions, emphasizing the need for tailored strategies (Hypotheses 2 and 3).

5. Discussion

5.1. Theoretical Implications

This study advances our understanding of how Agenda 2030 and the Paris Agreement shape corporate ESG performance, demonstrating how corporations adapt to evolving societal expectations and sustainability requirements. By analyzing corporate behavior through these policy lenses, the research contributes to the literature on corporate citizenship and CSR and aligns with ecological economics by connecting sustainability with resource allocation. The findings reinforce the global commitment to advancing sustainability goals and enrich contemporary frameworks for meeting the SDGs.

Our findings, particularly regarding the positive governance trajectory in developed economies, are consistent with recent systematic reviews, which emphasize the impact of robust governance and board diversity under contemporary sustainability frameworks like Agenda 2030. A “governance-first” pattern is consistent with SDG translation and international best practices, modulated by industry- and economy-level institutional capacities. These results also reinforce existing evidence that the regulatory context and international sustainability initiatives play a critical moderating role in ESG performance, as discussed in prior systematic literature reviews. Specifically, the governance findings reflect agencies’ and stakeholders’ theoretical predictions that stronger institutional and regulatory frameworks drive improved ESG outcomes—a well-documented association in systematic literature reviews such as that by Ed-Dafali, Adardour, Derj, Bami, and Hussainey [11]. Likewise, the observed variability in environmental and social performance between developed and developing economies, and between sectors, echoes the importance of contextual and legitimacy-related factors, further confirming conclusions from contemporary meta-analyses. No major divergences from these frameworks were observed in our sample, but sector-/country-specific patterns indicate fruitful directions for further research. The weaker and less consistent changes in SR performance may reflect greater heterogeneity in how “S” is measured across raters and regions, as well as slower policy convergence on social standards compared with environmental or governance dimensions. Prior studies highlight that social indicators remain less standardized and often rely on voluntary disclosure, weakening the overall trend (see Yusifzada et al. [102]). Future research should prioritize improved operationalization and cross-country comparability of the SR pillar. We recommend that future research prioritize standardized metrics for SR, harmonized global reporting practices, and enhanced data granularity to address persistent heterogeneity.

5.2. Managerial Implications

The findings of this study have several important implications for policymakers and corporate managers as they navigate the complexities of sustainability. First, it is evident that the positive trajectory in governance performance within developed economies—especially in the manufacturing sector—demonstrates the critical role of regulatory frameworks in driving ESG disclosure and shaping corporate strategies [37,76,103]. Governance enhancements are shown to positively influence corporate efficiency and financial outcomes, underscoring the value of transparent information disclosure [67,103]. Second, the observed improvements in environmental performance in developed countries reflect the effectiveness of environmental regulations in promoting green practices and sustainable innovation [71]. Environmental regulation is a critical external driver of green innovation in Chinese manufacturing, with command-and-control, market-based, and voluntary approaches each demonstrating significant positive impacts on firms’ innovation behavior [104]. Nevertheless, persistent challenges remain, particularly in meeting ambitious environmental targets. For instance, indecision regarding the United States’ participation in the Paris Agreement has impeded global environmental progress, as highlighted in the UN’s Emissions Gap Report [105]. Third, social performance continues to be difficult to regulate due to its voluntary nature, which points to the need for more consistent ESG disclosure practices and robust evaluation mechanisms [37,106]. Fourth, in developing economies, the underperformance in ESG metrics following the adoption of Agenda 2030 can be attributed to delayed regulatory implementation, as has been widely reported [50]. For example, China only introduced voluntary ESG reporting guidelines in 2022 [107], and other emerging economies such as Brazil, India, Russia, Mexico, Malaysia, and Thailand display varying degrees of regulatory maturity [108,109,110,111,112,113,114,115,116,117,118].

Recent declines in ESG performance in developing economies are better understood through country-specific regulatory timelines and the lens of institutional theory. For example, while China only introduced voluntary ESG guidelines in 2022, Brazil has partial mandatory disclosure, and India’s requirements remain fragmented, leading to substantial variance in ESG reporting and corporate adaptation [119,120]. Empirical research shows that regulatory milestones—such as the rollout of carbon trading in China and board mandates in Brazil—correlate directly with shifts in ESG trajectories, as timely policy enactment and enforcement influence how firms internalize sustainability standards [119].

Institutional theory further explains why “regulatory capacity” is critical for ESG adoption in developing economies: countries with limited institutional strength, fragmented enforcement, and weaker norms experience slower and less consistent uptake of sustainability practices [121,122]. These institutional gaps—manifesting as regulatory lag, ambiguous compliance requirements, and an absence of standardized reporting—limit the effectiveness of ESG initiatives and hinder cross-country convergence [122]. Our results align with these findings, suggesting that progress in ESG trajectories for developing economies is linked not only to formal regulatory milestones, but also to the broader institutional environment that shapes corporate responses.

Recent work supports the idea that ESG disclosure improves value, investor trust, and long-term reputation [123,124]. Our results align with the expectation that governance would lead to ESG improvements, reflecting swift policy-driven change and institutional compliance [48,49]. Environmental and social scores showed more conditional or delayed progress, paralleling the operational, regulatory, and contextual challenges that have been documented in recent empirical studies [50,51].

5.3. Policy Implications

The policy implications arising from this research are multifaceted. First, the positive trajectory in governance performance within developed economies, particularly in manufacturing, highlights the necessity of robust regulatory frameworks for improving corporate accountability. Policymakers should prioritize the development and enforcement of comprehensive ESG reporting regulations that encompass environmental, social, and governance disclosures. Calls for harmonized global frameworks and improved data quality in ESG reporting are increasing, aiming to standardize assessments and support regulatory compliance [103]. Clear and enforceable guidelines are essential for bridging governance gaps and aligning corporate practices with sustainability objectives [125]. Second, while the SDGs are aspirational, achieving them requires actionable policies that address fragmented and inconsistent sustainability practices. It is crucial that national policies are integrated with SDG frameworks to ensure effective implementation [125]. Furthermore, international collaboration is needed to harmonize ESG standards and establish unified reporting frameworks, thereby reducing fragmentation and facilitating global comparability [126]. Third, institutional integration is vital for fostering coordination among private, public, and international actors. Developing cross-sectoral policies, strengthening multistakeholder partnerships, and prioritizing capacity-building initiatives are pivotal, especially in developing economies. Tailored approaches that consider institutional complexity can support sustainability transitions in resource-constrained contexts [17]. Fourth, localized implementation of SDGs should incorporate participatory governance models that engage community organizations, subnational governments, and private stakeholders [126]. Last, to address declining environmental ratings in developing economies, capacity-building efforts must focus on creating governance systems and institutional mechanisms that support greener practices. Financial incentives, international cooperation, and robust environmental regulations are essential for achieving environmental targets. Education for sustainable development should emphasize capacity-building among policymakers and institutions to align sustainability efforts with local contexts [126]. The scaling up of regulatory intensity, alignment of policy instruments, and recognition of enterprise-led environmental responsibility have been vital in driving China’s shift from high-speed growth to high-quality, innovation-driven, and sustainable development [104,127]. Empirical evidence from Chinese listed firms shows that ESG performance—shaped in part by regulatory pressure and policy support, including fiscal subsidies and tax incentives—broadens the range of innovation resources and fosters higher-quality and more novel innovation outcomes [127]. The effectiveness of China’s policy toolkit—ranging from stricter enforcement of environmental regulations for heavy-pollution industries to the introduction of voluntary compliance initiatives—is well established and closely linked to measurable improvements in corporate green innovation performance [104]. Both of the above studies highlight the importance of improving and enforcing ESG evaluation systems and disclosure requirements to stimulate green innovation and support the transition towards sustainable economic growth [104,127]. These implications are in line with the recommendations from systematic literature reviews such as that by Ed-Dafali, Adardour, Derj, Bami, and Hussainey [11], which advocate for sector-specific governance reforms, enhanced board diversity, and integrated theoretical perspectives to inform policy and managerial decision-making in multinational ESG management.

5.4. Limitations and Future Research

This study is not without limitations. First, the absence of standardized ESG disclosure systems and limited data availability precluded the inclusion of small and medium enterprises (SMEs) in this longitudinal analysis. SMEs represent most firms and account for the largest proportion of global employment but face unique resource constraints, lower external pressure, and less rigorous regulatory environments in relation to ESG compared with large corporations. This often leads to more ad hoc and less formalized ESG adoption, where outcomes are shaped by capacity challenges and weaker stakeholder scrutiny. These major distinctions should be acknowledged when interpreting our findings. Future research should prioritize SME-specific data collection and comparative studies to clarify ESG performance and barriers beyond the large-cap cohort. This gap underscores the importance of examining SMEs’ contributions to sustainable development, particularly as they comprise 99% of firms and two-thirds of employment globally [128]. Examining SMEs using emerging scoring technologies and improved data collection methods could yield valuable insights into their sustainability practices. Second, although this study utilizes CSRHub as a comprehensive ESG data source, challenges regarding the consistency of ESG ratings remain, as highlighted by Berg, Kölbel, and Rigobon [46]. Divergences in ESG ratings—stemming from scope, measurement, and weight differences—complicate their application in assessing corporate performance. While CSRHub aggregates data from diverse sources to address these inconsistencies, future studies could benefit from using disaggregated data from individual rating agencies to refine analyses and account for potential biases, such as the rater effect. Last, future research should explore the dynamic interplay of sustainability transitions across industries and economies, addressing contextual challenges in both developed and developing regions. Expanding longitudinal studies to include smaller firms and alternative data sources could further enhance our understanding of the global sustainability agenda. Building on the benchmarking standards set by Ed-Dafali, Adardour, Derj, Bami, and Hussainey [11], this research underscores the need for further cross-country, multi-theory ESG studies to address the evolving complexities and practical demands identified in recent PRISMA-based reviews. This study did not incorporate placebo breakpoint or COVID-period robustness analyses due to sample constraints and the scope being focused on the SDG 2015 transition. Finally, we acknowledge that significant contemporaneous shocks—including the onset of the COVID-19 pandemic and the EU’s rapid ESG disclosure ramp-up after 2017—may confound clean attribution to the 2015 SDG break. Future research should isolate these effects using event-based or quasi-experimental variation and is encouraged to apply multi-break and pandemic-period robustness tests to further validate ESG trajectory changes. We also recommend that future studies develop standardized metrics for SR, utilize globally harmonized reporting frameworks, and focus on improved data availability at the social pillar level.

6. Conclusions

ESG scores have become critical metrics for evaluating corporate transparency, stakeholder engagement, and trustworthiness, and their role in assessing corporate performance is now indispensable. This study examines the impact of Agenda 2030’s global implementation on ESG performances across 320 Global Fortune 500 firms from 2010 to 2021 using a latent trajectory model.

First, the empirical findings reveal significant disparities between developed and developing economies. In developed economies, particularly within the manufacturing sector, ESG performance improved steadily from 2016 to 2021 following Agenda 2030’s adoption. These improvements highlight the effectiveness of international regulations, mandatory disclosure requirements, and customer-driven demands in shaping corporate sustainability practices. Second, developing economies exhibited declining ESG performance, largely due to the lack of robust national regulations [129]. Nonetheless, the promise of future international frameworks, enhanced information disclosure, and mounting customer pressures offer cause for optimism for sustainable progress in these regions. Third, the evolving landscape of ESG scoring demonstrates the financial outperformance of ESG-related initiatives. Regulatory and financial incentives, as well as increasing transparency, are driving this positive shift [100,130,131]. Regulatory measures, such as the CBAM and COP26’s goal of carbon neutrality by 2050, are expected to accelerate these trends. Future research should incorporate event study methodologies to evaluate the ongoing influence of these environmental regulations as global policies become more stringent, providing richer data for analysis.

In conclusion, ESG performance reflects the dual commitment of corporations to sustainable practices and the collective pursuit of the United Nations’ Sustainable Development Goals. This study underscores the necessity of clear regulations, standardized ESG practices, and collaborative global efforts. By prioritizing these actions, we can advance toward a more sustainable, equitable, and prosperous future, reinforcing the interconnected role of corporations, policymakers, and stakeholders in achieving these goals.