How does climate change asymmetrically affect economic policy uncertainty in the GCC countries: A multivariate quantile-on-quantile analysis

How does climate change asymmetrically affect economic policy uncertainty in the GCC countries: A multivariate quantile-on-quantile analysis

How does climate change asymmetrically affect economic policy uncertainty in the GCC countries: A multivariate quantile-on-quantile analysis

37 views
Categories: Articles
Authers: Mohamed Sami Ben Ali, Alanoud Al-Maadid, Brahim Bergougui
College of Business & Economics, Qatar University, Doha, Qatar
Received 2 May 2025, Revised 15 July 2025, Accepted 25 July 2025, Available online 28 July 2025, Version of Record 28 July 2025.

Abstract

This study investigates the critical relationship between climate change and EPU in GCC countries from 1980:M1 to 2023:M12, focusing specifically on how temperature changes influence EPU in these economies. The research distinguishes itself through its novel application of quantile-based methodologies, offering insights across different temporal periods and distribution quantiles. The findings reveal that GCC countries experience both benefits and vulnerabilities related to policy stability as temperatures change. Through M-QQR analysis, we find that Kuwait, Saudi Arabia, and the UAE maintain policy stability under lower temperature scenarios, while Qatar and Oman show diminished advantages. Crucially, all GCC countries exhibit increased policy uncertainty during extreme temperature events, highlighting the region's vulnerability to climate extremes. Our W-QR analysis adds a temporal dimension, showing that Saudi Arabia and Kuwait face immediate climate change impacts on policy uncertainty, reflecting their oil market sensitivities. Meanwhile, Oman, Qatar, and UAE exhibit delayed effects, with positive CC-EPU relationships emerging over medium to long-term periods. These findings underscore the importance of developing tailored policy responses that account for both the magnitude and timing of climate change impacts across different GCC economic structures, highlighting a critical asymmetry that demands attention from regional
As we progress through the new millennium, our planet faces an unprecedented environmental crisis. The Earth's climate system is undergoing rapid changes, with far-reaching consequences for all forms of life. Recent studies indicate a significant rise in global temperatures since the pre-industrial era [21,44]. Projections suggest a potential temperature increase of approximately 5 °C by 2100 if current trajectories persist [77]. This accelerated warming trend is not merely a matter of rising thermometer readings; it is reshaping the very fabric of our world. From food production to power generation, economic systems to societal structures, and from public health to geopolitics, no aspect of human existence remains untouched by this phenomenon [13,38,85]. The burden of this crisis, however, is not shouldered equally across the globe [97]. Economically disadvantaged countries, with limited access to cutting-edge technology and financial resources, bear the brunt of extreme weather events, often resulting in devastating economic setbacks [34,52]. Conversely, affluent countries situated in cooler regions are experiencing more rapid temperature increases, leading to substantial financial losses due to climate-related calamities [34]. One stark example is the European continent, which has incurred significant economic damages due to climate disasters, with losses estimated between 450 and 520 billion euros between 1980 and 2020 [37]. In response to this looming threat, countries worldwide have begun implementing various strategies at multiple levels of society. The Paris Agreement of 2015 established ambitious objectives to constrain global temperature rise to significantly below 2 °C above pre-industrial levels, with aspirations to limit this increase to 1.5 °C. The Intergovernmental Panel on Climate Change (IPCC) posits that achieving these targets necessitates a 45 % reduction in global greenhouse gas emissions by 2030 and the attainment of net-zero emissions by 2050 [63]. Yet, recent global events, such as armed conflicts in Eastern Europe, have introduced new layers of complexity to both climate-related and economic uncertainties. The resulting ambiguity in environmental policies, marked by a lack of clarity and consistency in governmental approaches to climate change (CC) mitigation and the shift towards sustainable energy sources, presents additional hurdles [43]. Given this context, understanding the magnitude and pathways of climate impacts on economic variables has become increasingly crucial for academics and policymakers alike [18,62,64].
Exogenous perturbations in economic systems may prompt governments to modify their policy stances, leading to biases in individuals' expectations regarding these policies. These biases, in turn, influence decision-making processes, culminating in what is termed economic policy uncertainty (EPU) [15]. In recent epochs, a combination of global shifts—including financial instability, migration crises, labor market imbalances, wealth inequality, demographic changes, and oil price fluctuations—has significantly increased the level of uncertainty in economic policy [39,74,90]. Empirical evidence indicates that elevated levels of EPU have a significant impact on macroeconomic variables, primarily by discouraging investment and consumption, which in turn slows economic growth and increases unemployment [56,95]. Moreover, EPU can also increase volatility in housing prices, potentially destabilizing real estate markets. Additionally, it may impact commodity markets through its influence on oil inventories and pricing mechanisms [12,59].
Given these far-reaching implications, it is imperative to elucidate the determinants of EPU. Climate change emerges as a significant factor in this context, exerting influence on diverse economic variables such as agricultural yields, public health outcomes, consumer spending patterns, and operational expenses. This, in turn, engenders macroeconomic instability [1,30]. In response to climate-related risks, governments frequently formulate and recalibrate economic policies aimed at ameliorating adverse impacts on the real economy. However, this adaptive process inadvertently amplify EPU [11,92]. This phenomenon is particularly pronounced in emerging economies, such as those of the Gulf Cooperation Council (GCC) countries - Bahrain, Kuwait, Oman, Qatar, Saudi Arabia, and the United Arab Emirates -. In these developing countries, climate change has the potential to significantly amplify EPU. It achieves this by disrupting established economic patterns and introducing considerable unpredictability into various sectors of economic activity.
Understanding the climate-EPU relationship requires recognition of its inherently asymmetric nature, which manifests through multiple dimensions of heterogeneity. The asymmetric impact of climate change on economic policy uncertainty encompasses three distinct but interconnected forms of asymmetry that this study systematically examines. First, distributional asymmetry refers to the differential effects of climate change across varying levels of economic policy uncertainty. This form of asymmetry recognizes that climate impacts do not uniformly affect all uncertainty scenarios; rather, extreme climate conditions may disproportionately amplify high-uncertainty states compared to low-uncertainty conditions. Second, temporal asymmetry captures the divergent patterns of climate-EPU relationships across different time horizons. This dimension acknowledges that immediate policy responses to climate shocks may differ substantially from long-term structural adjustments, creating distinct short-term volatility patterns versus chronic uncertainty trajectories. Third, country-specific asymmetry reflects the heterogeneous responses across nations based on their unique economic structures, institutional capacities, and climate vulnerabilities. This form of asymmetry recognizes that countries with different economic foundations, policy frameworks, and environmental exposures will exhibit varying degrees of sensitivity to climate-induced policy uncertainty. The comprehensive examination of these three asymmetric dimensions provides a nuanced understanding of how climate change creates differential impacts on economic policy uncertainty, moving beyond traditional linear assumptions to capture the complex, nonlinear relationships that characterize climate-economy interactions.
The temporal evolution of economic policy uncertainty across GCC economies demonstrates heterogeneous responses to exogenous shocks and structural transformations over five decades. Fig. 1 illustrates EPU trajectories for five GCC countries (1970–2024), revealing the transition from stable hydrocarbon-dependent systems to increasingly volatile policy environments. Initial periods exhibited minimal EPU levels, consistent with the economic stability characteristic of early oil exploitation phases. However, country-specific divergences emerged following major economic and geopolitical disruptions, including the Gulf War (1990–1991), global financial crisis (2008–2009), and oil price volatility episodes. The UAE demonstrated pronounced EPU spikes during 1990 (0.55), 1998, 2012, and 2021, while Kuwait exhibited notable fluctuations in 1972 and 1991. Saudi Arabia displayed persistent upward trends with significant volatility, particularly during 1991 and 2008, reflecting economic diversification challenges. Qatar maintained relative stability compared to regional peers, while Oman experienced delayed EPU increases post-2009, coinciding with global financial instability and commodity price fluctuations.
Fig 1
  1. Download: Download high-res image (665KB)
  2. Download: Download full-size image

Fig. 1. Evolution of EPU trends in each of the five GCC countries (1970–2024).

Source: WUI [98].
The nexus between climate change and economic policy uncertainty becomes empirically evident through systematic temperature analysis across the GCC region. Fig. 2 presents temperature trajectories for five GCC countries (1970–2024), revealing consistent warming trends aligned with global climate patterns. The data demonstrates accelerated warming post-1990, coinciding with intensified industrialization and urbanization processes. Saudi Arabia exhibits the steepest temperature increase, particularly pronounced after 1990, aligning with rapid economic transformation and urban development. The UAE shows marked temperature elevation from the late 1980s, corresponding to oil-driven economic expansion and extensive infrastructure development. Kuwait displays similar warming patterns from the early 1990s, with heightened vulnerability due to its compact geography and fossil fuel dependence. Oman demonstrates moderate warming trends post-1995, while Qatar shows accelerating temperature increases from the late 1980s, consistent with its energy-intensive economic structure and high per capita emissions.
Fig 2
  1. Download: Download high-res image (557KB)
  2. Download: Download full-size image

Fig. 2. Evolution of average temperature trends in each of the five GCC countries (1970–2024).

Source: Gortan et al. [45]
The academic literature on EPU has evolved along two distinct paths. The first examines how EPU shapes both macroeconomic and microeconomic outcomes [46,65,94]. The second investigates the multifaceted factors that shape EPU itself [74,78].While this dual approach has enriched our understanding of EPU, there remains a notable gap in understanding how climate change specifically affects EPU in GCC countries. The juxtaposition of Fig. 1Fig. 2 reveals intriguing patterns that motivate our research. The warming trends shown in Fig. 2, particularly the acceleration after 1990, coincide with periods of heightened EPU volatility illustrated in Fig. 1. This temporal alignment suggests a potential relationship between rising temperatures and economic policy uncertainty, particularly during periods of extreme climate events or sustained temperature increases. This study aims to address this gap by analyzing the CC-EPU relationship in the GCC context. GCC economies are particularly vulnerable to environmental changes due to their geographical location and economic structures, which are heavily reliant on hydrocarbon exports. The rising temperature trends demonstrated in Fig. 2, combined with the volatile EPU patterns shown in Fig. 1, underscore the critical need to understand how climate change asymmetrically affects economic policy uncertainty in these strategically important economies.
Based on these discussions, the main value added of this paper can be generalized in the following aspects:
  • First, this study represents a pioneering contribution to the literature by examining the asymmetric relationship between climate change and EPU. While previous research has extensively explored climate change impacts on traditional economic indicators such as GDP growth, inflation, and employment, this investigation addresses a significant gap by focusing on policy uncertainty as a critical yet underexplored outcome variable. The research significance lies in its recognition that climate shocks create complex challenges for policymakers, potentially leading to increased uncertainty in economic policy formulation and implementation. This relationship is particularly relevant given the growing frequency and intensity of extreme weather events, which can disrupt economic activities and complicate policy responses. By establishing empirical evidence for the climate-EPU nexus, this study provides valuable insights for understanding how environmental risks translate into economic policy challenges. The cross-country comparative approach enhances the robustness of findings while offering practical guidance for policymakers seeking to navigate the intricate relationship between climate variability and economic stability. This contribution is essential for developing more resilient economic policies in an era of increasing climate uncertainty.
  • Second, the study's methodological innovation lies in its integration of two sophisticated quantile-based analytical techniques, creating a comprehensive framework for examining complex economic relationships. The Multivariate Quantile-on-Quantile Regression (M-QQR) technique enables the examination of how different intensities of climate change influence varying levels of economic policy uncertainty across the entire distribution. Unlike conventional regression methods that focus on average effects, M-QQR captures the heterogeneous impacts across different quantiles, revealing whether extreme climate events disproportionately affect high-uncertainty scenarios compared to low-uncertainty conditions. This distributional analysis provides nuanced insights that traditional linear models cannot capture, offering a more complete understanding of the climate-EPU relationship. Complementing this approach, the Wavelet Quantile Regression (W-QR) method introduces a temporal dimension to the analysis, decomposing the climate-EPU relationship across multiple time horizons. W-QR distinguishes between immediate policy responses to climate shocks and long-term structural adjustments, providing insights into both short-term volatility and chronic uncertainty patterns. This temporal decomposition is crucial for understanding how climate impacts evolve over different time scales, enabling more targeted policy interventions. The combination of M-QQR and W-QR creates a comprehensive analytical framework that addresses both distributional heterogeneity and temporal dynamics. This dual approach provides a more complete picture of the climate-EPU relationship, capturing both the intensity of effects across different scenarios and their evolution over time. This methodological innovation establishes a replicable framework that can be applied to other complex economic relationships.
  • Third, the research delivers substantial practical value through its policy-relevant insights, bridging the gap between academic research and real-world decision-making. By quantifying the relationship between climate variability and economic policy uncertainty, the study provides policymakers with empirical evidence to anticipate and prepare for climate-induced economic disruptions. This evidence base is particularly valuable for developing proactive rather than reactive policy responses, enabling more effective crisis management and economic stabilization efforts. The findings facilitate the design of adaptive economic strategies that address both environmental and economic risks simultaneously. Understanding how climate change amplifies policy uncertainty enables the development of diversification strategies, infrastructure investments, and institutional reforms that enhance overall economic resilience. This practical relevance makes the research directly applicable to policy planning processes and strategic decision-making frameworks.
  • Fourth, the study's focus on Gulf Cooperation Council (GCC) countries provides valuable insights into a region with unique environmental and economic characteristics. The GCC region's combination of arid climate conditions and hydrocarbon-dependent economies creates a compelling case study for climate-economy interactions. The region faces intensifying climate risks including rising temperatures and water scarcity, while simultaneously pursuing economic diversification initiatives that introduce additional policy complexities. This context makes the region particularly suitable for examining climate-EPU relationships. The cross-country examination within the GCC framework accounts for variations in economic structures, policy priorities, and climate vulnerabilities across member states. This granular approach ensures that findings are relevant to individual country contexts while maintaining regional coherence. The analysis supports country-specific policy recommendations while contributing to broader understanding of climate-economy interactions in resource-dependent economies. The research directly supports ongoing regional initiatives such as Saudi Vision 2030 and similar diversification programs across GCC countries. By providing evidence on how climate factors influence policy uncertainty, the study informs strategic planning processes and supports the development of climate-resilient economic policies tailored to regional conditions.
The paper proceeds as follows. Section 2 reviews and synthesizes the literature on climate change and economic policy uncertainty. Section 3 describes our research design and methodology, and presents the empirical results alongside robustness checks. Section 4 interprets and discusses these findings. Finally, Section 5 concludes by highlighting the study’s key insights.

2. Climate change and EPU: a theoretical overview

The theoretical framework explaining the relationship between climate change and EPU is rooted in policy uncertainty theory, as developed by Baker et al. [15], which posits that policy unpredictability significantly affects economic decision-making across businesses, governments, and individuals. This relationship manifests through four key theoretical channels:
  • First, climate change creates direct economic disruptions through physical damage to assets and human capital, affecting production, investment, and consumption patterns. The climate variability caused by phenomena like El Niño1 and La Niña2 impacts food security, supply chains, and energy stability [28,41], compelling governments to implement reactive policies that increase short-term EPU [32]. This dynamic is further shaped by broader institutional factors, where differing priorities and public sentiment influence policy decisions, often favoring immediate economic concerns over environmental objectives [42].
  • Second, global carbon reduction goals necessitate substantial economic policy adjustments, particularly affecting high-emission industries. The implementation of climate change mitigation policies requires significant investments in green technology and renewable energy, creating new dimensions of EPU as countries navigate the complex coordination of these transitions [89]. This process involves intricate trade-offs between immediate economic gains and long-term sustainability goals [10], affecting fiscal and industrial policies [15].
  • Third, behavioral economics principles developed by Tversky & Kahneman [53,86] suggest that under conditions of high climate change and EPU, economic actors exhibit cognitive biases such as loss aversion and status quo bias. These behavioral patterns can impede necessary adaptations in energy use, consumer behavior, and sustainable technology investment, potentially weakening global cooperation on climate change mitigation as domestic economic issues take precedence over international environmental commitments [10].
  • Fourth, the transition risks associated with climate change create uneven impacts across industries, particularly affecting sectors dependent on high energy and resources use. The emergence of stranded assets, increased operational costs, and labor market disruptions presents significant challenges for policymakers [75]. This is especially pronounced in economies reliant on resource-intensive industries, where the development of new economic drivers often lags behind policy expectations [61], creating additional EPU as governments attempt to balance employment, economic growth, and carbon emission reduction goals [96].
This theoretical framework, building on Bloom's [26] work, emphasizes that the relationship between climate change and EPU creates a complex feedback loop, where environmental imperatives and economic stability concerns continuously interact, requiring innovative policy approaches that can effectively address both climate risks and economic uncertainty while supporting sustainable transitions.

3. Empirical framework

3.1. Data and variables

By considering both data availability and sample representativeness, this paper focuses on five GCC economies —Kuwait, Oman, Qatar, Saudi Arabia, and the United Arab Emirates—as the subjects of investigation.3 The study examines the effects of climate change on economic policy uncertainty (EPU) within the region, utilizing data spanning from M1–1970 to M9–2024. Due to gaps in the climate change data, interpolation techniques were employed to create a continuous time series for analysis.

3.1.1. Dependent variable -economic policy uncertainty (EPU)

In this research, the World Uncertainty Index (WUI) serves as a proxy indicator for economic policy uncertainty (EPU). The WUI quantifies the prevalence of uncertainty-related terminology within Economic Intelligence Unit (EIU) publications. The computation of the WUI adheres to a systematic methodology comprising several stages:
  • Information Source: The foundation of the WUI lies in the periodic EIU reports, which offer comprehensive global economic assessments and projections at the national level.
  • Content Examination: These EIU reports undergo computational linguistic analysis to detect and enumerate instances of ambiguity-related lexicon, including variations of the root term.
  • Standardization: To address disparities in document length, the raw frequency is calibrated against the total word count of each report.
  • Consolidation: The adjusted figures are then amalgamated to yield a singular WUI metric for each temporal or geographic unit of analysis.
  • Statistical adjustment: To mitigate short-term fluctuations and present a more stable representation, the WUI undergoes a mathematical transformation using a rolling mean over three consecutive periods.
Elevated WUI scores signify heightened uncertainty of economic policy, as evidenced by the increased occurrence of uncertainty-related terminology in EIU analyses. For this research, WUI metrics were acquired from the comprehensive repository at worlduncertaintyindex.com, which maintains an extensive collection of WUI data encompassing numerous countries and geographical areas.

3.1.2. Independent variable - climate change

This study employs climate change as the core independent variable, operationalized through temperature data as a proxy measure. The research utilizes average temperatures (in Celsius) from Weighted Climate Dataset built by Gortan et al. [45],4 applying a grid area weighting method to calculate monthly temperatures and aggregate them to the national level. This approach is chosen due to the uncertainty regarding whether human economic factors respond more significantly to temperature changes than non-human factors [9,52].
Our selection of temperature as a proxy for climate change is based on three key factors:
  • Prevalence in economic literature: Temperature is the most commonly used variable to represent climate change in economics research.
  • Correlation with other climate indicators: Temperature increases are strongly correlated with various climate-related phenomena, including extreme weather events, droughts, storms, sea level rise, and carbon dioxide emissions.
  • Data availability and richness: Annual mean temperature data is widely available across various datasets (common weather, gridded, and reanalysis), offering extensive temporal and geographical coverage.

3.1.3. Control variables

Building on extensive literature on the determinants of EPU [79,83], this study incorporates control variables—exchange rate, inflation, FDI, and trade openness—to capture key transmission channels of EPU, particularly in trade and fiscal policy contexts. Trade policies influence exchange rates, trade volumes, and FDI flows, while fiscal policies shape inflation, exchange rates, and the broader economic environment. To address missing monthly data for GCC countries, interpolation and extrapolation from annual data were employed. Transforming annual data to monthly frequencies using the quadratic match-average method [19,20,22,24] mitigated data limitations by reducing fluctuations and isolating non-seasonal components, enhancing reliability. Economic indicators and temperature data were log-transformed to address magnitude disparities. Table 1 provides detailed descriptions of each variable.

Table 1. Summary of data.

CodeIndicator nameMeasurementSource
CCClimate Change (Ln)Average temperature measured in degrees CelsiusGortan et al. [45]
EPUEconomic policy uncertaintyIndex calculated based on the frequency of uncertainty-related terms in EIU reportsWUI (2024)
ERExchange Rate (Ln)Official exchange rate (local currency units per US dollar, period average)WDI (2024)
TOTrade openness (Ln)Sum of exports and imports of goods and services as a percentage of GDPWDI (2024)
FDIForeign Direct Investment (Ln)Net inflows of foreign direct investment as a percentage of GDPWDI (2024)
INFInflation (Ln)Annual percentage change in the GDP deflatorWDI (2024)
Note. WDI: World Development Indicators. WUI: World Uncertainty Index. EIU: Economic Intelligence Unit.

3.2. Multivariate quantile-on-quantile regression (M-QQR)

The impact of CC on EPU can be complicated by various factors, potentially leading to biased results when using traditional bivariate quantile-quantile regression (B-QQR) due to omitted variable bias [14]. Drawing from the studies of [3,7,8,[68][69][70],87,88], we utilize the novel M-QQR approach to mitigate this issue. The M-QQR methodology represents an advancement over the B-QQR by incorporating the moderating influence of other exogenous variables through interactions, allowing for a more nuanced and accurate assessment of the relationship under study. This method enables a more comprehensive analysis by simultaneously accounting for multiple factors. Specifically, it evaluates the impact of CC on EPU while controlling for factors such as exchange rate, inflation, FDI, and trade openness in the five GCC countries. The relationship is parameterized assuming that EPU follows a quantile distribution, with the -quantile represented as a function of CC and other unidentified factors. We define the relationships as:
Where EPU represents economic policy uncertainty, CC stands for climate change, and U refers to the n undefined matrices. Incorporating the control variables into Eq. (1), the equation can be reformulated as follows:
Since the function is unknown due to the undefined moderating effects, we explore the dependence between the γ-quantile of EPU and the τ-quantile of climate change. The first-order Taylor expansion of  is:
Eq. (4) demonstrates that the functions  and  are indexed by both γ and τ, while  is indexed solely by τ. Additionally, the first-order Taylor expansion of  shows that both  and  are similarly indexed by γ and τ.  and can be alternatively expressed as  and which are embedded functions of  and . Replacing these into Eq. (3) results in:
The Eq. (5) can be represented as  except for , which signifies the  quantile of the EPU movement, and conditional on the τth quantile of CC. In this case,  and  are indexed by γ and τ. and the full tail correlation is captured by (*). This (*) part of the equation also considers uncertainty factors, leading to Eq. (5), which can be rewritten as:
This equation can also be denoted as (*) except for . The optimization problem for this equation is given by:
Here, , and  denote the estimated values for , and , respectively. To address this optimization problem, a solution for  is necessary. Considering the localized effect of the τ quantile of CC, a Gaussian kernel K(·) is applied to determine the weight of  within its domain, with bandwidth h set accordingly. The difference between  and  is represented by:

3.3. Wavelet quantile regression (W-QR)

As a further analysis to complement our main investigation of climate change's impact on EPU using M-QQR, we employed an additional methodological approach to capture the multifaceted nature of this relationship across different time scales and quantiles. Recent literature has highlighted that associations between variables often exhibit variations across distinct time scales [2,4]. To account for this complexity, we adopted the Wavelet Quantile Regression (W-QR) method introduced by [6]. This approach allows us to investigate the impact of climate change on EPU across various time scales and quantiles, addressing some of the limitations inherent in traditional QR techniques. Following the methodology outlined by Kumar and Padakandla [57], we decomposed the time series data of our dependent variable (EPU) and independent variable (climate change) using the Maximal Overlap Discrete Wavelet Transform (MODWT) developed by Percival and Walden [73]. This decomposition enables us to analyze the relationship between climate change and EPU at different frequency bands, providing insights into how this relationship may vary over short-term, medium-term, and long-term horizons. The W-QR method also aligns closely with recent advancements in climate-economy research, particularly as demonstrated in Adebayo and Özkan [6], who applied similar wavelet-based techniques to investigate nonlinear and time-varying interactions between environmental and economic variables. By combining the insights from M-QQR and W-QR, we offer a dual-perspective investigation —one that captures both the distributional heterogeneity and the multi-temporal complexity of the CC-EPU relationship. This hybrid analytical strategy not only enhances the robustness of our findings but also contributes to the growing toolkit of advanced econometric methods in climate policy research.

3.4. Preliminary empirical assessments

This section of the study presents the results of our empirical preliminary analysis. We follow a structured approach, with each step of our analysis corresponding to the sequence outlined in Fig. 3. This figure serves as a visual roadmap, guiding readers through our analytical process.
Fig 3
  1. Download: Download high-res image (344KB)
  2. Download: Download full-size image
Fig. 3
Before moving on to the main analysis, we perform some preliminary analysis on the log-monthly data series.

3.4.1. Summary statistics

The descriptive statistics for monthly EPU and temperature data in GCC countries (Table A2, Appendix) reveal negative skewness and kurtosis in temperature data, indicating occasional low extremes and flatter distributions. The Jarque-Bera test confirms non-normality in both EPU and temperature data, with the UAE's EPU data showing the largest deviation. These characteristics, coupled with observed variations in EPU and temperature patterns, underscore the suitability of nonlinear techniques, such as quantile regressions, for analyzing their relationships in the GCC region.

3.4.2. BDS linearity test

As previously mentioned, descriptive statistics suggest that the study variables may exhibit nonlinear characteristics. To verify this, we apply the BDS linearity test, developed by [27] and used in studies such as [19,23,25]. The results of the BDS test, presented in Table 2, confirm that all the study variables display nonlinear behavior over the sample period. These nonlinear traits affirm the suitability of using the M-QQR method in this study.

Table 2. Brock, Dechert, Scheinkman (BSD) test outcomes.

CountryVariableDimension 2Dimension 3Dimension 4Dimension 5Dimension 6Decision
KuwaitES0.00000.00000.00000.00000.0000Nonlinear
AI0.00000.00000.00000.00000.0000Nonlinear
OmanES0.00000.00000.00000.00000.0000Nonlinear
AI0.00000.00000.00000.00000.0000Nonlinear
QatarES0.00000.00000.00000.00000.0000Nonlinear
AI0.00000.00000.00000.00000.0000Nonlinear
Saudi ArabiaES0.00000.00000.00000.00000.0000Nonlinear
AI0.00000.00000.00000.00000.0000Nonlinear
United Arab EmiratesES0.00000.00000.00000.00000.0000Nonlinear
AI0.00000.00000.00000.00000.0000Nonlinear
Note. Values indicate p-values.

3.4.3. Quantile unit root results

In this study, unit root testing, a critical step in time-series analysis to ensure data stationarity, is conducted using a novel quantile-based approach. Following Adebayo and Özkan [5] and Liu et al. [60], the Quantile Augmented Dickey-Fuller (Q-ADF) and Quantile Phillips-Perron (Q-PP) tests are employed, offering advantages over conventional methods by capturing distributional variations across quantiles. The methodology involves generating quantile series for the variables at intervals of 0.05 from 0.05 to 0.95, as per [58], and applying ADF and PP tests to these series. Results, visualized in Fig. 4Fig. 5, confirm the stationarity of the logarithmic series across the entire conditional distribution, validating the data for further analysis without transformations. This quantile-based approach provides a nuanced understanding of stationarity that traditional tests may overlook.
Fig 4
  1. Download: Download high-res image (411KB)
  2. Download: Download full-size image
Fig. 4
Fig 5
  1. Download: Download high-res image (393KB)
  2. Download: Download full-size image
Fig. 5

3.5. Main estimation results

3.5.1. M-QQR results

In the next phase of analysis, we apply the nonparametric M-QQR methodology to examine the non-linear relationship between CC (average temperature) and EPU, controlling for key macroeconomic variables—exchange rate, inflation, FDI, and trade openness—within each GCC economy. The M-QQR slope coefficients, depicted in Fig. 6, capture the influence of the τth quantile of CC on the λth quantile of EPU, considering the moderating effects of these variables.
Fig 6
  1. Download: Download high-res image (1MB)
  2. Download: Download full-size image
Fig. 6
According to Fig. 6, In Kuwait, climate change, as measured by average temperatures, influences EPU across different quantiles in varying ways. At lower quantiles of climate change (0.05–0.35), there is a significant curbing effect on EPU, suggesting that modest temperature increases initially help stabilize economic policy environments. However, this beneficial relationship deteriorates as temperatures rise to middle quantiles (0.40–0.60), where the effect reverses, leading to increased EPU. In higher temperature quantiles (0.65–0.95), climate change exerts an even stronger impact, amplifying EPU. This pattern indicates that Kuwait’s economic policy framework is relatively resilient under lower temperatures but becomes increasingly vulnerable as temperatures escalate, suggesting a threshold effect in the climate change–EPU relationship. Thus, while milder temperatures initially help reduce EPU, the stabilizing effect weakens and ultimately reverses with intensified climate change, heightening economic policy uncertainty at higher temperature levels.
In Oman, throughout the low to middle temperature quantiles (0.05–0.80 °C), climate change shows negligible impact on EPU, suggesting a robust policy environment that remains relatively stable under moderate temperature variations. However, the relationship changes dramatically at higher temperature quantiles (0.80–0.95), where increasing temperatures correspond to a significant positive effect on EPU. This indicates that while Oman's economic policy framework shows resilience under moderate temperature conditions, it becomes particularly vulnerable to uncertainty when temperatures reach their highest levels, pointing to potential adaptation challenges in extreme temperature scenarios.
Qatar demonstrates a unique three-phase relationship between CC and EPU. In the lower to middle temperature quantiles (0.05–0.70), CC shows minimal influence on EPU, suggesting policy resilience under moderate temperature conditions. The relationship then shifts notably in the upper-middle quantiles (0.75–0.80), where temperature increases actually help reduce EPU, indicating a temporary stabilizing effect. However, at the highest temperature quantiles (0.85–0.95), this beneficial effect reverses sharply, leading to increased EPU. This pattern suggests that Qatar's economic policy framework has a narrow optimal temperature range where it can effectively manage climate-related uncertainties.
Saudi Arabia reveals a pattern similar to Kuwait's, but with distinct thresholds in its impact on EPU. At lower temperature quantiles (0.05–0.35), CC effectively reduces EPU, indicating that modest temperature increases can initially stabilize economic policy. However, this beneficial effect reverses at the middle quantiles (0.40–0.65), where rising temperatures begin to amplify EPU, suggesting a threshold where temperature increases shift from stabilizing to destabilizing economic policy. At higher quantiles (0.70–0.95), this relationship becomes even stronger, indicating that Saudi Arabia's economic policy framework faces growing challenges in maintaining stability under elevated temperature conditions. This pattern underscores the critical role of temperature thresholds in influencing economic policy responses to climate change in Saudi Arabia.
The UAE shows a distinctive three-tier relationship between CC and EPU. At the lowest temperature quantiles (0.05–0.20), CC has a negative impact on EPU, helping stabilize economic policy in milder climate conditions. This effect becomes neutral across low to middle quantiles (0.25–0.60), suggesting a zone of policy resilience. However, at higher temperature quantiles (0.75–0.95), the relationship transforms dramatically, showing a strong positive correlation between temperature increases and EPU. This pattern reveals that the UAE's economic policy framework is most effective at managing climate-related uncertainties only under very specific temperature conditions, with increasing vulnerability at higher temperatures.
In summary, the comprehensive M-QQR analysis across these GCC countries reveals important regional patterns in the CC-EPU relationship. While Kuwait, Saudi Arabia, and the UAE experience beneficial effects in terms of economic policy stability at low temperature quantiles, Qatar and Oman demonstrate different response patterns. A common thread across all five countries is the increased economic policy uncertainty associated with higher temperature quantiles, suggesting a regional vulnerability to extreme temperature conditions.

3.5.2. W-QR results

The W-QR analysis, depicted in the heatmaps of Fig. 7, reveals a distinct temporal pattern in the relationship between CC and EPU for Kuwait and Saudi Arabia, which are heavily reliant on oil exports and possess substantial oil reserves. In the short and medium term, CC exhibits a negative influence on EPU across the quantiles. This indicates that initial, moderate temperature increases may actually contribute to stabilizing economic policy, likely due to the consistent revenue generated by oil exports. However, as the time horizon extends, this effect turns positive, with rising temperatures correlating with increased EPU across all quantiles. This evolving dynamic implies that the influence of CC on EPU in Kuwait and Saudi Arabia is not static, but rather changes over time. The two oil-dependent economies are increasingly vulnerable to the long-term risks of climate change. As global calls for sustainability and reduced fossil fuel reliance grow, these countries face heightened uncertainty in their economic policies. Their significant oil production activities make them particularly sensitive to climate variations and shifts in global energy markets. Climate-related regulations can further amplify policy uncertainty over time, posing challenges to these economies heavily reliant on fossil fuels.
Fig 7
  1. Download: Download high-res image (549KB)
  2. Download: Download full-size image
Fig. 7
The W-QR analysis for Oman, Qatar, and the United Arab Emirates paints a distinct picture compared to Kuwait and Saudi Arabia. In the short term, the impact of CC on EPU is negative, albeit categorized as 'weak' across all quantiles. Oman, with smaller resource reserves, and Qatar and the UAE, with more diversified economies, experience only a modest stabilizing effect from minor temperature changes. However, in the medium and long term, CC’s influence turns positive across quantiles, potentially heightening EPU as global shifts toward sustainability impact these economies differently. For Qatar, where natural gas plays a central role, and for the UAE’s mixed economy, such climate-related pressures may prompt greater shifts in policy direction. Oman, with its early diversification efforts, could experience unique responses as it balances limited resources and adaptation strategies.
These W-QR findings, visually represented in heatmaps, underscore how each GCC nation’s resource dependence shapes the time-sensitive impacts of CC on EPU. In Saudi Arabia and Kuwait, climate change’s effects on EPU reflect the vulnerabilities of heavily oil-reliant economies, whereas Oman, Qatar, and the UAE—benefiting from more diversified economies—show a different temporal pattern. Here, the positive CC-EPU relationship appears in the medium to long term, suggesting that economic diversification may offer more adaptability in managing climate-related policy uncertainties. However, even the more diversified GCC countries are not immune to the challenges of reconciling climate change mitigation with economic growth and policy stability as they continue to evolve, highlighting the need for tailored economic policy approaches in response to climate change pressures across the GCC.

3.6. Robustness checks

To verify the accuracy of our m-QQR analysis, we conducted two distinct validation tests.

3.6.1. M-QR results

Our first validation approach employed the established M-QR method, which is frequently used to validate m-QQR results in current research. While single-variable QR could be used, we specifically chose m-QR to match our model structure. Fig. 8 compares the results of both methods across different EPU quantiles, and the similar patterns observed in both M-QR and QQR results support the validity of our findings.
Fig 8
  1. Download: Download high-res image (398KB)
  2. Download: Download full-size image
Fig. 8

3.6.2. KRLS results

Our study employed kernel-based regularized least squares (KRLS) as a secondary validation method - a sophisticated machine learning approach developed by Hainmueller and Hazlett [47] and further enhanced by Ferwerda et al. [40]. The KRLS approach was selected for its ability to detect non-linear relationships in the data without requiring assumptions of linearity [71]. Specifically, we applied KRLS to calculate pointwise marginal effects of climate change on EPU across individual data points for GCC economy. The results, visualized in Fig. 9, demonstrate consistency with our M-QQR analysis, thus reinforcing the robustness of our primary findings.
Fig 9
  1. Download: Download high-res image (663KB)
  2. Download: Download full-size image
Fig. 9

3.6.3. Addressing endogeneity issues

Estimating the causal effect of climate change (proxied by temperature) on EPU presents significant endogeneity challenges, primarily arising from reverse causality, omitted variable bias, and potential measurement error [55]. While rising temperatures may increase global uncertainty and elevate EPU, socioeconomic conditions and policy responses influenced by EPU can also affect future temperature trajectories. This bidirectional relationship creates a feedback loop, complicating causal inference. To address these issues, we employ the Instrumental Variable Quantile Regression (IVQR) framework [54]. Our baseline specification instruments for contemporaneous temperature change using lagged temperature levels. Historical temperatures capture inherent climatic persistence and slow-moving environmental processes, making them strong predictors of current temperature changes. Crucially, lagged values are exogenous to contemporaneous EPU distributions, effectively mitigating simultaneity and reverse causality concerns [97]. Although rising temperatures can amplify global uncertainty and increase EPU—potentially creating bidirectional causality—the lagged instrument isolates temperature variation driven by historical conditions, uncontaminated by concurrent EPU fluctuations. This approach provides a robust identification strategy, yielding reliable estimates of the causal impact of temperature change (our climate change proxy) on EPU after correcting for endogeneity.
Fig. 10 presents the IVQR coefficient estimates, which exhibit patterns consistent with the MQQR results shown in Fig. 6 and align closely with the findings in Fig. 10. This consistency reinforces the robustness of our results. For instance, in Qatar, the impact of climate change on EPU is negative at lower quantiles but turns positive at higher quantiles—a trend that is consistently captured by both methods (IVQR and MQQR).
Fig 10
  1. Download: Download high-res image (336KB)
  2. Download: Download full-size image
Fig. 10

3.6.4. Alternative climate indicator

To address concerns regarding the multidimensional nature of climate change and ensure our findings are not driven by the choice of climate proxy, we conducted robustness checks using precipitation as an alternative climate indicator. While temperature serves as our primary proxy due to its direct relevance to the GCC region's extreme heat conditions and its established relationship with economic outcomes in hot climates [29], we acknowledge that climate change encompasses multiple dimensions including precipitation patterns, extreme weather events, and climate volatility. Our robustness analysis using precipitation data largely confirms our main findings.
Fig. 11 presents the MQQR results with precipitation as the climate variable, which demonstrate remarkably consistent patterns with our temperature-based estimates shown in Fig. 6. The quantile-specific relationships between climate conditions and EPU remain fundamentally unchanged across most countries and quantiles, providing confidence in the robustness of our core findings. However, some nuanced differences emerge that merit discussion. Most notably, for Oman, the effect at the 60th percentile shifts from negative with temperature to positive with precipitation, while effects at other quantiles remain consistent. This divergence may reflect Oman's unique geographic position and exposure to monsoon patterns, where precipitation variability creates distinct policy uncertainties compared to temperature fluctuations. Despite this isolated difference, the overall quantile-based patterns and cross-country heterogeneity identified in our main analysis remain robust to the choice of climate indicator. These results suggest that while the specific magnitudes of climate-EPU relationships may vary slightly depending on the climate dimension considered, the fundamental nonlinear patterns and country-specific heterogeneity we document are not artifacts of our temperature focus but reflect genuine structural relationships between climate conditions and economic policy uncertainty in the GCC region.
Fig 11
  1. Download: Download high-res image (1MB)
  2. Download: Download full-size image
Fig. 11

4. Discussion and policy implications

The examination of climate change’s influence on EPU across the GCC reveals rich, context‑specific insights: the region’s economic structures, degrees of climate resilience, and reliance on natural resources critically modulate policy responses. Our combined M‑QQR and W‑QR results demonstrate that climate impacts on EPU vary not only across temperature quantiles but also over different time horizons—short, medium, and long—highlighting divergent vulnerabilities and adaptive capacities. Interpreted through both distributional and temporal lenses, these findings uncover complex, nonlinear relationships that both align with and extend prior research. As illustrated in Fig. 12, the climate‑EPU linkage exhibits distinct quantile‑based profiles that shift according to each country’s economic and climatic characteristics.
Fig 12
  1. Download: Download high-res image (963KB)
  2. Download: Download full-size image
Fig. 12
From a quantile perspective, Kuwait, Saudi Arabia, and the UAE benefit from economic stability at lower temperature quantiles, indicating that their policy frameworks can effectively manage normal temperature variations. This resilience is consistent with Dell et al. [35] and Hsiang and Jina [48], who show that countries with stronger institutional structures tend to handle moderate climate impacts more effectively. However, as temperatures rise to higher quantiles, the results reveal increased EPU across all GCC countries. This vulnerability aligns with Nordhaus [66] and Deryugina [36], which detail how higher temperatures can destabilize even well-prepared economies, a finding further corroborated by Burke et al. [29] who emphasize the challenges of managing extreme weather events, particularly in already hot regions. This is also aligning with Weitzman [93] on the concept of “tail risks” in climate change, where extreme events create heightened economic instability. The GCC region’s vulnerability to these tail risks is underscored by Pal and Eltahir’s [72] research, which suggests that parts of the region could become uninhabitable by 2100 if extreme climate events intensify. However, important heterogeneity emerges across individual countries, reflecting their distinct economic structures and policy approaches. Kuwait demonstrates remarkable resilience at lower temperature quantiles, likely attributable to its substantial sovereign wealth fund and diversified investment portfolio, which provides a buffer against climate-induced uncertainty. Saudi Arabia exhibits similar patterns, benefiting from its Vision 2030 diversification strategy and massive infrastructure investments in renewable energy projects like NEOM, which may enhance confidence in the country's long-term climate adaptation capabilities. The UAE's response pattern reflects its advanced economic diversification, with Dubai's service-oriented economy and Abu Dhabi's strategic investments in renewable energy creating institutional resilience. The country's early adoption of green finance initiatives and its hosting of COP28 demonstrate proactive climate policy engagement, potentially explaining its relative stability at moderate temperature levels.
Qatar and Oman present distinctly different patterns that deviate from regional trends, warranting closer examination. Qatar's three-phase pattern—where moderate temperatures offer stability, but extreme conditions increase EPU—reflects the country's unique economic structure. As a small, gas-rich nation with limited economic diversification despite recent efforts, Qatar's economy remains vulnerable to extreme climate events that could disrupt its energy infrastructure and threaten its comparative advantage in LNG exports. The country's heavy reliance on desalination and air conditioning makes it particularly sensitive to temperature extremes, as highlighted by Kahn et al. [52] in their analysis of adaptation costs in hot climates. This supports Tol [84]'s findings on diversification as a climate resilience factor, suggesting that Qatar's concentrated economic base amplifies uncertainty under extreme conditions. Oman's response pattern is particularly intriguing, showing relative stability under low to moderate temperatures but facing distinct challenges under extreme conditions. This pattern reflects Oman's early and sustained diversification efforts, including its development of manufacturing, tourism, and logistics sectors, which have reduced its dependence on oil revenues compared to other GCC states. The country's Oman Vision 2040 emphasizes sustainable development and climate resilience, potentially explaining its superior performance at moderate temperature levels. However, Oman's geographic exposure to cyclones and its limited fiscal resources compared to wealthier GCC neighbors may explain its vulnerability to extreme temperature events, as these could strain its adaptation capacity and fiscal buffers. These country-specific patterns highlight the critical importance of economic diversification, institutional quality, and fiscal capacity in determining climate resilience. While the GCC region shares common vulnerabilities due to its hot climate and energy-dependent economies, the heterogeneous responses underscore that one-size-fits-all policy recommendations may be inappropriate. Countries with greater diversification and stronger institutions demonstrate enhanced resilience, while those with concentrated economic structures face amplified uncertainty under extreme conditions.
The temporal analysis (W-QR) reveals a clear divergence among GCC countries. In the short term, Kuwait and Saudi Arabia experience a stabilizing effect from moderate temperatures, likely benefiting from sustained oil revenues. However, as the time horizon extends, CC begins to positively influence EPU, signaling heightened uncertainty. This rapid, time-sensitive response is in line with findings by de Brauw et al. [33] on immediate climate impacts and Batten et al. [16], who explore how energy-dependent economies respond quickly to climate shocks. This immediate response also underscores the specific vulnerabilities of oil-heavy economies that are acutely sensitive to both physical climate risks and global energy shifts. In contrast, Oman, Qatar, and the UAE show a gradual increase in EPU over the medium to long term. This delayed response aligns with Su et al. [82] and Stern [80]’s insights on the cumulative impact of climate change, where gradual risks accrue, impacting policies over time. The concept of “tragedy of the horizon” introduced by Carney [31] is particularly relevant here, as climate risks that seem minimal in the short term may create substantial long-term challenges. For the UAE and Qatar, whose economies are more diversified, medium- to long-term climate impacts signify potential adaptation needs, especially as these countries balance sustainability with continued growth.
The distinct responses across quantiles and time horizons have significant policy implications, particularly for economies heavily reliant on fossil fuels. The need for immediate policy responses in oil-dependent countries like Kuwait and Saudi Arabia corresponds to Sanfilippo et al. [76] and IRENA [51] analysis of the energy transition challenges faced by petroleum-exporting countries. As EPU rises over the medium and long term in diversified economies like the UAE and Qatar, Stern et al. [81] emphasize the importance of structural economic transformation, which would allow these countries to adapt more effectively to sustained climate pressures.
The overall regional vulnerability pattern, especially at higher temperature quantiles, corresponds to the [50] projections for the Middle East, which indicate escalating policy challenges in response to climate change. The IMF [49] has also pointed to the intersection of climate vulnerability and economic transition needs in the GCC, reinforcing that even diversified economies within the region cannot entirely escape the effects of climate change. The temporal differences in climate-EPU relationships, with oil-dependent economies facing immediate impacts and diversified ones showing delayed responses, align with findings from the World Bank [91] that indicate economic diversification as a moderating factor in climate resilience.
This study advances the literature on climate change and economic policy by illustrating the complex, multi-layered nature of the CC-EPU relationship across temperature distributions and time horizons. The findings underscore the need for nuanced, differentiated policy approaches across the GCC, recognizing that each country’s economic structure shapes its specific climate risks. Consistent with Battiston et al. [17] on climate risk management, tailored policy frameworks will be essential, addressing both immediate shocks in oil-heavy economies and the need for long-term structural adjustments in diversified ones. Given the GCC’s position at the intersection of climate vulnerability and energy transition pressures [67], policy frameworks must balance the dual objectives of managing immediate climate shocks and planning for long-term resilience. This includes integrating climate risk into economic policy and actively investing in green technologies, infrastructure, and renewable energy sources. These policies can help the GCC manage increasing climate-related uncertainties while supporting a more sustainable economic transition.

5. Conclusions

Historically, GCC countries prioritized oil market stability and economic diversification over climate action when addressing global challenges. However, recent years have witnessed a paradigm shift, with climate change mitigation becoming central to the region's economic transformation strategies. This evolution is driven by multiple factors: global energy transition, rising temperatures, regional security concerns, and international pressure for carbon reduction commitments. As GCC countries navigate oil market volatility, regional geopolitical tensions, and the need for economic diversification, they increasingly recognize that climate action and economic stability are complementary rather than competing priorities. This integrated approach is exemplified in initiatives like Saudi Arabia's Vision 2030, UAE's Net Zero 2050 strategic initiative, and Qatar's National Environment and Climate Change Strategy, reflecting a new era where economic resilience and climate adaptation are viewed as interconnected imperatives.
Building on this evolving policy landscape, this study investigates the critical relationship between climate change and economic policy uncertainty in GCC countries from 1980m1 to 2023m12, focusing specifically on how temperature changes influence policy uncertainty in these economies. The research distinguishes itself through its novel application of quantile-based methodologies, offering insights across different temporal periods and distribution quantiles—an approach previously underutilized in climate change and EPU research. The study employs sophisticated statistical frameworks, primarily the M-QQR and W-QR, to capture the inherent nonlinearities in the dataset. To ensure result robustness, the study incorporates M-QR and KRLS estimators as complementary verification tools.
The findings from our analysis paint a nuanced picture of climate change impacts across GCC countries. The M-QQR analysis reveals that Kuwait, Saudi Arabia, and the UAE experience policy stability benefits at lower temperature quantiles, while Qatar and Oman show distinct response patterns. Notably, all GCC countries demonstrate heightened policy uncertainty at higher temperature quantiles, indicating regional vulnerability to extreme temperatures. Our W-QR analysis adds a temporal dimension, showing that Saudi Arabia and Kuwait face immediate climate change impacts on policy uncertainty, reflecting their oil market sensitivities. Meanwhile, Oman, Qatar, and UAE exhibit delayed effects, with positive CC-EPU relationships emerging over medium to long-term periods. Policymakers in oil-dependent economies like Kuwait and Saudi Arabia should prioritize immediate climate adaptation strategies, while diversified economies like the UAE and Qatar need long-term structural reforms to enhance resilience. Investments in green technologies, renewable energy, and climate-resilient infrastructure are essential to mitigate risks and support sustainable transitions. Tailored approaches are crucial to balance short-term vulnerabilities with long-term resilience across GCC countries, aligning with their unique economic structures and climate risks.
Despite these significant contributions, the study acknowledges several limitations that provide avenues for future research.
  • First, the analysis exclusively focuses on GCC economies, which, while providing deep insights into resource-rich, climate-vulnerable nations, limits the generalizability of findings to other economic contexts. Expanding the scope to examine the relationship between EPU and climate factors across diverse economic blocs such as BRICS nations, OECD countries, or other regional groupings like ASEAN would substantially enhance the understanding of these dynamics across different developmental stages, economic structures, and climate vulnerabilities. Such comparative analyses could reveal whether the observed patterns are unique to oil-dependent economies or represent broader global phenomena.
  • Second, the study's reliance on temperature as the primary climate indicator, while justified by data availability and consistency, may not capture the full spectrum of climate-related risks that influence economic policy uncertainty. Future studies could incorporate additional climate metrics including precipitation variability, extreme weather event frequency, drought indices, and composite climate risk measures to provide a more holistic assessment of climate impacts on policy uncertainty.
  • Third, the temporal scope of the analysis, while spanning significant climate and economic events, could be extended to capture longer-term climate cycles and their policy implications. Additionally, the use of the World Uncertainty Index as the sole EPU proxy, though well-established, could be complemented by alternative uncertainty measures such as policy-specific uncertainty indices, market-based uncertainty indicators, or text-based policy uncertainty measures to enhance robustness.
  • Finally, the study's focus on aggregate national-level relationships could be enriched by sector-specific analyses that examine how different economic sectors within GCC countries respond to climate variations in terms of policy uncertainty. This sectoral approach could reveal heterogeneous effects across industries such as energy, agriculture, manufacturing, and services, providing more targeted policy insights.
By addressing these limitations, forthcoming studies could further enrich our knowledge of the intricate interactions between climate change and economic policy uncertainty, contributing valuable insights to this rapidly evolving field and informing more nuanced policy responses to climate-related economic challenges.

Funding

Research reported in this publication was supported by the Qatar Research Development and Innovation Council under grant #ARG01-0508-230093. The content is solely the responsibility of the authors and does not necessarily represent the official views of Qatar Research Development and Innovation Council.

CRediT authorship contribution statement

Mohamed Sami Ben Ali: Validation, Supervision, Resources, Project administration, Methodology, Investigation, Funding acquisition. Alanoud Al-Maadid: Validation, Supervision, Resources, Project administration, Methodology, Investigation, Funding acquisition. Brahim Bergougui: Writing – original draft, Visualization, Software, Investigation, Formal analysis, Data curation, Conceptualization.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Data availability

Data will be made available on request.


Keywords:
Climate change