Assessing the Environmental Impact of Fiscal Consolidation in OECD Countries: Evidence from the Panel QARDL Approach

Abstract

Concerns about ensuring a sustainable environment are growing, attracting major attention from policy professionals worldwide. Therefore, this study investigates the nonlinear impacts of fiscal consolidation on CO₂ emissions in 17 OECD countries from 1978 to 2020. To probe the short- and long-term connections across various quantiles of CO₂ emissions, we adopted panel QARDL frameworks. The Granger non-causality test was used to investigate the variables’ association with CO₂ emission. The study’s main findings confirm the overall beneficial effect of fiscal consolidation on carbon emissions. It reduces CO₂ emissions at almost all quantiles in the short run. By contrast, in the long run, the effect is positive at lower quantiles and turns negative at upper quantiles. Furthermore, a causality analysis identified a bidirectional causal relationship between fiscal consolidation and CO₂ emissions, confirming the existence of mutual influence. While Keynesian theory links fiscal consolidation to economic recession, our findings support the non-Keynesian view, showing that such policy can foster economic growth and thereby contribute to reducing CO₂ emissions in the short run. Thus, OECD countries are orienting public spending and carbon taxation toward environmentally friendly practices while ensuring environmental protection and deficit reduction. Nonetheless, the identified mixed effect in the long run highlights the need for sustained consolidation policies by enhancing expenditure efficiency and adopting targeted taxation measures to achieve lasting emission reductions and support the transition to cleaner energy, even when emissions are relatively low.

1. Introduction

Environmental deterioration is a major challenge for many countries today, posing the most serious risk to global wealth and sustainability. Based on the Intergovernmental Panel on Climate Change’s (2022) report, if no effective measures are implemented to reduce greenhouse gas emissions, global temperature is expected to increase to 1.5 °C by 2030–2050. It is widely suggested that global warming and climate change present a serious threat to humanity’s existence, economic survival, and economic development progress. According to Landrigan et al. (2018), inefficiencies caused by pollution-related illnesses can cause yearly GDP reductions of about 2% in low- to middle-income countries. Our contemporary most critical concern is preserving a sustainable climate while maintaining economic development (Dincer, 2000; Moldan et al., 2012). Energy is thereby deeply ingrained in all facets of economic, social, and environmental advancement. Furthermore, after the 2008 financial crisis, several countries sought to implement effective macroeconomic policies, particularly fiscal consolidation programs, to speed up their economic recovery. This had an impact on numerous macroeconomic factors. Accordingly, a growing body of research reveals that fiscal policies play an important role in determining environmental degradation (Lopez et al., 2011; Halkos & Paizanos, 2013; Galinato & Islam, 2014; Lopez & Palacios, 2014; Islam & Lopez, 2015).

Considering the above-mentioned gaps, economists are looking more closely at how different factors affect carbon emissions. In the climate change context, fiscal policies, which heavily influence aggregate demand, economic expansion, income distribution, and inflation control, have become a growing focus of interest. Many economies have moved towards fiscal consolidation, aiming to safeguard fiscal sustainability and reduce public deficit (Alesina et al., 2015). Following the so-called non-Keynesian approach, fiscal consolidation occurs through spending cuts or tax increases, generating expectations about future credible fiscal consolidation and future income improvement (Giavazzi & Pagano, 1990; Blanchard, 1990; Sutherland, 1997). This approach includes neoclassical elements, such as the Ricardian Equivalence property. This implies that wealth and expectational effects might have an important role in the improvement of economic activity, stimulating economic growth. In the aftermath of the 2008 global financial crisis, a large number of countries incorporated fiscal consolidation programs in their fiscal laws to reduce their debt and deficit for sustaining growth. Empirical studies on developed countries, especially OECD countries, have proven that they experience an expansionary effect, particularly an increase in consumption, investment and thus economic growth, when implementing fiscal consolidation programs (Alesina & Ardagna, 1998, 2010, 2012; Blanchard, 1990; Alesina & Perotti, 1995, 1997; Giudice et al., 2007; McDermott & Wescott, 1996; Lambertini & Tavares, 2005; Afonso et al., 2006; Alesina et al., 2019).

Fiscal consolidation through spending cuts and tax increases is closely linked to environmental indicators because it influences both the level and the structure of public budgets. On the expenditure side, consolidation may reduce discretionary expenditures. On the revenue side, it often involves changes in taxation, including potential increases in ecotaxes. Both mechanisms play a role in stimulating growth and reallocating resources devoted to environmental protection and green investment that affect incentives for cleaner production and consumption. These mechanisms justify our focus on the fiscal consolidation–emissions nexus.

In OECD countries, fiscal consolidation has been the main strategy for addressing rising public debt, especially after 2008 financial crisis. While much attention has been given to the macroeconomic implications of fiscal consolidation, its potential impact on climate change remains largely unexplored. This gap in research presents an important opportunity to investigate how these contractionary fiscal policies, also referred to fiscal consolidation, might influence environmental outcomes. Understanding this interaction is essential for evaluating whether debt reduction strategies are compatible with long-term sustainability goals. This study therefore aims to fill this gap by examining the effect of fiscal consolidation on CO₂ emissions in OECD countries.

Considering the contribution of the current research, we could not discover any notable studies on the chosen analyses in the OECD countries. With fiscal consolidation and climate policy as high-priority policy issues, gaining a deeper understanding of this relationship appears essential. Moreover, by 2050, OECD economies are projected to generate nearly 40% of global GDP, a shift that is likely to significantly accelerate global energy demand (OECD, 2015). As a result, knowing the consequences of rigorous fiscal policies on renewable energy could be fruitful even for other developed economies. Taking the preceding context into account, this study aims to investigate the relationship between fiscal consolidation and carbon dioxide (CO₂) emissions using annual data from 1978 to 2020. A fiscal consolidation episode is identified when a government undertakes specific initiatives by increasing taxes and reducing spending to lower its fiscal deficit. Instead of using the primary balance to identify these episodes, researchers have turned to the cyclically adjusted primary balance (CAPB), which adjusts for the effects of economic fluctuations and provides a clearer picture of intentional policy actions. While CAPB is a useful measure, recent studies recommend the narrative approach proposed by Romer and Romer (2010), Ramey and Shapiro (1998), and Ramey (2011), which attempts to create “exogenous” measures of fiscal policy. According to these authors, the changes in CAPB might still reflect economic fluctuations because its adjustments rely on estimates of potential output, leading to biased results. Thus, the narrative approach has greater accuracy in reflecting policymakers’ true intentions to limit fiscal deficits, which minimizes issues of misclassification and endogeneity. To the best of our knowledge, our study is the first to examine the impact of fiscal consolidation episodes on CO₂ emissions in OECD countries, using newly constructed data based on the narrative method from Adler et al. (2024), which effectively overcomes the limitations of traditional approaches. This method draws on a historical perspective by relying on the direct analysis of budget documents, policy statements, and other official records. Additionally, the study’s selection of the 1978–2020 sample period is crucial as it includes major economic and environmental policy shifts in OECD countries. As a result, it reflects a wide array of scenarios, contributing to a detailed exploration of long-term trends. Lastly, another key contribution of this study is its use of the advanced panel Quantile Autoregressive Distributed Lag (QARDL) approach by Cho et al. (2015), rather than relying on conventional methods to estimate the impact of fiscal consolidation on pollution. The main advantage of using such an asymmetric estimation model is to observe the long-term effects and related short-term movements of fiscal consolidation episodes on OECD’s CO₂ emissions on different quantiles of CO₂ emissions. These insights equip policymakers with a clearer understanding of how to design effective plans and policies for environmental preservation.

The remaining sections of this study are organized as follows: Section 2 depicts the literature review. Section 3 illustrates the empirical model, variable description, and data source. Section 4 discusses empirical results. Section 5 concludes this paper and presents useful policy implications.

2. Literature Review

This research identifies fiscal policy as a critical component influencing the demand side of the economy. Fiscal policy instruments, particularly taxes and government expenditures, are closely linked to GDP growth, production capacity, and energy consumption (Muhafidin, 2020). Consequently, government expenditure and revenue collection directly and indirectly affect total energy consumption and environmental quality. Numerous recent studies have drawn attention to the effect of fiscal policy on the environment and claimed it is increasingly oriented toward achieving environmental sustainability, especially in countries facing high levels of climate change, including OECD countries. Some studies underscore the important contribution of fiscal policy in maintaining low levels of carbon emissions (Katircioglu & Katircioglu, 2018; Halkos & Paizanos, 2016; Kamal et al., 2021; Burke, 2019), while others emphasize its potential to deteriorate environmental quality (Yuelan et al., 2019).

The impacts of contractionary and expansionary shocks of fiscal policy instruments on CO₂ emissions have been a subject of conflicting findings. The existing literature surrounding this issue has largely focused on expansionary measures, with relatively few studies examining the impact of contractionary fiscal policies on CO₂ emissions.

The literature on expansionary fiscal consolidation and its impact on CO₂ emissions is extensive. Generally, empirical evidence indicates that the impact of fiscal policy on CO₂ emissions differs across countries based on their development level, fiscal policy orientation, and environmental goals. For instance, Lopez et al. (2011) underscore the critical role of fiscal spending composition and empirically analyze its effects on the environment. These authors demonstrate that pollution is more likely to decrease when government expenditure is reallocated to social and public purposes. Halkos and Paizanos (2016) used quarterly data from 1973 to 2013 to investigate the effects of fiscal spending on CO₂ emissions in the US. The authors highlighted that the impact of fiscal policy varies depending on the type of pollution, the fiscal policy approach, and the changes in the composition of government spending. Katircioglu and Katircioglu (2018) examined the case of Turkey using data from 1960 to 2013 and concluded that increasing public spending reduces carbon emissions. They confirmed that fiscal policies targeting the energy sector have been effective in Turkey. According to Halkos and Paizanos (2013), a negative relationship between government spending and CO₂ emissions is observed in low-income countries; however, this relationship shifts to positive with rising income levels. They propose combining government spending cuts with suitable environmental standards and establishing international environmental agreements. Yuelan et al. (2019) investigated the influence of public revenue and spending in China from 1980 to 2016, concluding that fiscal policy expansion is harmful to the environment due to a lack of innovative and sustainable manufacturing practices. Mahmood et al. (2022) discovered that government spending boosts CO₂ emissions in Gulf Cooperation Council economies. Their findings suggest that fiscal policy prevents production- and consumption-based, pollution-producing economic activity. The findings demonstrate that although oil continues to be a substantial source of exports and income, it also contributes significantly to rising pollution. Abbass et al. (2022) demonstrated that government spending may promote sustainability initiatives, resulting in lower CO₂ emissions in Pakistan. Thus, fiscal policies aiming at lowering emissions are worthwhile.

Research investigating the effects of contractionary fiscal policy on emissions remains limited. For example, Huynh (2020) investigated the fiscal policy effect on pollution in developing Asian nations. They demonstrate that higher government spending contributes to lower air pollution levels, while tax hikes have the opposite effect, suggesting the rise in funding for environmental initiatives to support a more environmentally friendly economy. Chishti et al. (2021) found that contractionary fiscal policy reduces CO₂ emissions in the BRICS countries. Their estimation results indicate that such a policy may limit excessive energy use and economically damaging operations. Wolde-Rufael and Mulat-Weldemeskel (2021) examined carbon taxes in a panel of seven emerging economies and discovered an inverted U-shaped association with CO₂ emissions, implying that the benefits of environmental policies emerge over time.

To summarize the overall findings of the above literature, fiscal consolidation, through government spending cuts and tax increase, is expected to boost disposable income and consumption, pushing industries to expand production, leading to higher CO₂ emissions. Conversely, contractionary fiscal policy, by raising taxes and cutting spending, reduces aggregate demand, stimulating firms to scale back production and fossil fuel use, thereby lowering carbon emissions. A significant research gap remains, particularly in testing the effect of fiscal consolidation from the so-called non-Keynesian perspective. According to this theory, under certain conditions, fiscal contraction can lead to economic growth rather than recession, as lower public deficits and debt levels may boost private sector confidence (Mtibaa et al., 2022). Several investigations, including Giavazzi and Pagano (1996), Alesina and Perotti (1997), Alesina and Ardagna (1998, 2010), and Giavazzi et al. (2000), demonstrated that fiscal consolidations are more likely to have a non-Keynesian effect, promoting GDP growth. These authors confirm that fiscal consolidation can enhance private sector expectations and perceived wealth, lower real interest rates, and consequently stimulate consumption and investment.

Due to escalating pollution and emission levels, sustainable development and environmental protection have emerged as global imperatives in recent years. Developed or advanced economies such as OECD countries appear to prioritize environmental issues and allocate funds to address them. The OECD, which intensively applies fiscal consolidation policy to reduce its deficit, is a good case study to explore the impact of public budget on energy sectors. Fiscal consolidation is a policy aimed at reducing deficits and stimulating economic growth, thereby generating the budgetary capacity needed to support eco-friendly investment. In OECD countries, successful consolidation is generally designed to cut unproductive expenditures while protecting or prioritizing spending that promotes environmental sustainability. On the revenue side, the use of environmental taxes plays a dual role: it supports fiscal balance and at the same time provides resources that can be reinvested in technological innovation, renewable energy, and climate policies. In this way, fiscal consolidation combines its spending and revenue components to align sound public finances with sustainable environmental outcomes. Over the last two decades, many OECD nations have intensified their environmental policies by adopting carbon pricing mechanisms, emissions trading schemes, and renewable energy mandates. Based on International Energy Agency data, five countries, France, Germany, Japan, the United Kingdom, and the United States, members of the OECD, contributed to 66% of the total public energy spending on energy research in 2021 (IEA, 2021a). From 2016 to 2020, the collective public spending on R&D for OECD countries had almost 38% growth (IEA, 2021b). Furthermore, R&D strategic plans such as the Strategic Energy Technology Plan employed by the EU in 2007 indicate that most OECD countries have had progressing climate policies with a focus on energy innovation (Garrone & Grilli, 2010). Although the literature offers extensive empirical evidence on the relationship between environmental policies and sustainability outcomes, much of it focuses on emerging economies such as China. More recent studies on OECD countries, however, highlight a direct connection between budgetary allocations and environmental performance. For instance, Umar and Safi (2023), for instance, use movement quantile regression (MMQR) analysis to test the significance of green finance and green innovation in achieving sustainable development. Their findings highlight that green finance and public spendings on innovation contribute significantly to reducing CO₂ emissions in the OECD. According to these authors, budgetary measures targeted toward green investment can have a measurable effect on environmental quality. Ziaei (2025) tested the relationship between public spending on R&D for renewable and non-renewable energy and CO₂ emissions. A Panel-NARDL model was used in 10 OECD countries for the period 1990–2020. The results show that an increase in public expenditure on the R&D of energy efficiency can induce economic activity and enhance CO₂ emissions. Thus, CO₂ emissions in OECD countries could be reduced by applying the public budget to promote R&D in energy sectors. Additional evidence from OECD economies comes from Sohag et al. (2024), who examine the effect of environmental policy stringency (EPS) on CO₂ emissions. Their results suggest that stricter policies reduce emissions once certain thresholds are reached, although the effect diminishes in highly industrialized economies. Importantly, they emphasize the need for complementary budgetary measures such as innovation subsidies to maximize the effectiveness of EPS.

Given that OECD countries have adopted fiscal consolidation as a key policy strategy to restore public finances, understanding its implications for CO₂ emissions is crucial. If fiscal contraction fosters economic growth, its environmental impact might differ from the traditional Keynesian view, which suggests that reduced economic activity typically associated with fiscal contraction leads to lower CO₂ emissions. This contributes to the debate on fiscal consolidation’s environmental impact, especially in countries where it is a key policy, and supports the aim of this study to assess this relationship empirically.

3. Data and Methodology

3.1. Data

This study explores the environmental impact of fiscal consolidation using a range of explanatory variables. Fiscal consolidation (FC) is measured through action-based consolidation efforts, primarily driven by the objective of reducing budget deficits. We investigate whether such consolidation policies have a significant effect on environmental outcomes, particularly CO₂ emissions per capita (CO₂). We rely on the database compiled by Adler et al. (2024)1, which extends the work of Alesina et al. (2018)2 and Devries et al. (2011), while maintaining a narrative-based approach. Their construction involves analyzing historical budget documents, IMF reports, and other official sources to identify discretionary fiscal actions not driven by the economic cycle. Based on this approach, their dataset includes 242 fiscal consolidation episodes across 17 OECD countries, Australia, Austria, Belgium, Canada, Denmark, Finland, France, Germany, Ireland, Italy, Japan, the Netherlands, Portugal, Spain, Sweden, the United Kingdom, and the United States, from 1978 to 2020. Some years with negative total fiscal adjustment appear in the dataset as deficit-driven fiscal consolidations because they are part of broader multi-year consolidation processes. To identify genuine fiscal consolidation episodes, we exclude such cases, retaining only years with positive total adjustment. Consequently, this study uses the FC variable that captures the provided positive total fiscal consolidation values as the sum of the discretionary tax increases and spending cuts, measured in percentage of GDP. This variable takes the value of zero when no fiscal consolidation occurs. GDP per capita (GDP) reflects the overall economic performance of a country. Foreign direct investment (FDI) is measured as net inflows as a percentage of GDP and captures international investment activity. Urban population (URB) refers to the percentage of the total population living in urban areas. Before displaying the whole panel QARDL model illustration, we added the country indicator (i) and time indicator (t) as subscripts:

$${CO}_{2_{it}}={\mathsf{\beta }}_0+{\mathsf{\beta }}_1{FC}_{it}+{\mathsf{\beta }}_2{FDI}_{it}+{\mathsf{\beta }}_3{URB}_{it}+{\mathsf{\beta }}_4{GDP}_{it}+{\epsilon }_{it}$$

(1)

where ${\mathsf{\beta }}_1,{\mathsf{\beta }}_2,{\mathsf{\beta }}_3$ and ${\mathsf{\beta }}_4$ are the loading parameters of FC, FDI, URB, and GDP, respectively. ${\epsilon }_{it}$ is an error term. The data are collected from the World Development Indicators (WDI). The fiscal consolidation episodes, however, are identified using newly constructed data for OECD countries provided by Adler et al. (2024).

Table 1. Results of descriptive statistics.

	Obs	Mean	Min.	Max.	Std. Dev.	Skewness	Kurtosis	Jarque–Bera (J-B)	p-Value
CO2	765	10.0699	2.4873	22.3303	4.1664	0.8141	3.0035	84.4987	0.000
FC	765	0.3369	0.04	5.23	0.7573	3.0963	14.276	5275.194	0.000
URB	765	76.5920	41.979	98.153	11.1542	−0.5501	3.1974	39.8337	0.000
FDI	765	3.3544	−31.3058	86.4791	7.9071	4.6835	38.4389	42,829.14	0.000
GDP	765	286.5861	9.01599	4473.208	888.9702	3.877348	16.5836	7798.2589	0.000

presents the descriptive statistics of the variables. All the mean values are positive. While CO₂, FC, FDI, and GDP exhibit positive skewness, URB is found to have negative skewness. GDP has the largest recorded standard deviation (888.9702), whereas FC has the smallest (0.7573). the Jarque–Bera statistic rejects the null of normality at the 1% level for all variables. This supports the adoption of the panel QARDL technique by confirming that the chosen data are not normally distributed. The QARDL model captures asymmetric and nonlinear relationships across different quantiles, which may reflect underlying structural changes. Figure 1 depicts the dynamics of GDP, FC, CO₂, FDI, and URB over the sample period.

3.2. Methodology

Several investigations in the previous literature illustrate that although financial and economic aspects are important in determining environmental sustainability, these particular effects can change significantly according to circumstances. To further identify these different impacts, the QARDL model is employed. This model extends the traditional ARDL approach by capturing both short- and long-term impacts across different quantiles. This makes it possible to evaluate CO₂ emissions at various levels of CO₂ emissions within OECD nations more sophisticatedly. Indeed, such model does not estimate a single “average” effect, for all countries. Instead, it shows how the impact of fiscal consolidation on CO₂ emissions can change depending on the country’s pollution level. Thus, the effect could vary depending on a country’s existing level of pollution, providing a more nuanced understanding of heterogeneous impacts. In this regard, the effect of fiscal consolidation may be different in low-emission countries than in high-emission ones. This helps capture variations that a standard ARDL model would overlook. The panel QARDL model has the advantage of enhancing non-normal errors by considering the skewness, heterogeneity, and outliers of the dependent variable. Another advantage of the QARDL technique is that it can handle short-term changing dynamics. Additionally, the QARDL method exhibits outstanding efficiency compared with other asymmetrical techniques since it employs a “data-driven” procedure to produce nonlinearity in the series instead of the partial sum methodology used by Shin et al. (2014) for the NARDL method.

Getting more specific, the panel ARDL model utilized in this study is formulated as follows:

$$\begin{gathered} {CO}_{2_{it}}=\mu +{\sum }_{\mathrm{j}=1}^{\mathrm{p}}{\sigma }_{{CO}_{2_i}}{CO}_{2_{it-j}}+{\sum }_{\mathrm{j}=0}^{q_1}{\sigma }_{{FC}_i}{FC}_{it-j}+{\sum }_{\mathrm{j}=0}^{q_2}{\sigma }_{{FDI}_i}{FDI}_{it-j}+{\sum }_{\mathrm{j}=0}^{q_3}{\sigma }_{{URB}_i}{URB}_{it-j}+ \\ {\sum }_{\mathrm{j}=0}^{q_4}{\sigma }_{{GDP}_i}{GDP}_{it-j}+{\epsilon }_{it} \end{gathered}$$

(2)

where ${\epsilon }_{it}$ is described as ${CO}_{2_{it}}-E\left[{CO}_{2_{it}}/F_{it-1}\right]$ with $F_{\mathrm{i}\mathrm{t}-1}$ is the smallest field made by (${FC}_{it},{FDI}_{it},{URB}_{it},{GDP}_{it},{FC}_{it-1},{FDI}_{it-1},{URB}_{it-1},{GDP}_{it-1}$). p, $q_1$, $q_2$, $q_3$, and $q_4$ refer to lag orders relying on the Schwarz information criterion (SIC). Using Cho et al.’s (2015) method, it is required to modify the fundamental Equation (1) in the quantile panel ARDL form:

$$\begin{gathered} {\mathrm{Q}}_{{CO}_{2_{it}}}=\mu (\mathsf{\tau })+{\sum }_{\mathrm{j}=1}^{\mathrm{p}}{\sigma }_{{CO}_{2_i}}\left(\tau \right){CO}_{2_{it-j}}+{\sum }_{\mathrm{j}=0}^{q_1}{\sigma }_{{FC}_i}{\left(\tau \right)FC}_{it-j}+ \\ {\sum }_{\mathrm{j}=0}^{q_2}{\sigma }_{{FDI}_i}\left(\tau \right){FDI}_{it-j}+{\sum }_{\mathrm{j}=0}^{q_3}{\sigma }_{{URB}_i}\left(\tau \right){URB}_{it-j}+{\sum }_{\mathrm{j}=0}^{q_4}{\sigma }_{{GDP}_i}\left(\tau \right){GDP}_{it-j}+{\epsilon }_{it}(\mathsf{\tau }) \end{gathered}$$

(3)

where ${\epsilon }_{it}\left(\mathsf{\tau }\right)={CO}_{2_{it}}-{CO}_{2_{it}}(\mathsf{\tau }/{\mathrm{F}}_{\mathrm{i}\mathrm{t}-1})$ and $\mathrm{Q}{CO}_{2_{it}}(\mathsf{\tau }/{\mathrm{F}}_{\mathrm{i}\mathrm{t}-1})$, and 0 < τ < 1 denotes the level of quantile. Given that Equation (3) has a serial correlation term, the panel QARDL model may be written as

$$\begin{gathered} {\mathrm{Q}}_{\mathsf{\Delta }{CO}_{2_{it}}}=\mu +\mathsf{\rho }{CO}_{2_{it-1}}+{\mathsf{\delta }}_{\mathrm{C}\mathrm{F}}{FC}_{it-1}+{\mathsf{\delta }}_{\mathrm{F}\mathrm{D}\mathrm{I}}{FDI}_{it-1}+{\mathsf{\delta }}_{\mathrm{U}\mathrm{R}\mathrm{B}}{URB}_{it-1}+{\mathsf{\delta }}_{\mathrm{G}\mathrm{D}\mathrm{P}}{GDP}_{it-1}+ \\ {\sum }_{\mathrm{j}=1}^{\mathrm{p}-1}{\Phi }_{{co2}_i}{\Delta co2}_{it-j}+{\sum }_{\mathrm{j}=0}^{q_1}{\Phi }_{{FC}_i}{\Delta FC}_{it-j}+{\sum }_{\mathrm{j}=0}^{q_2}{\Phi }_{{FDI}_i}{\Delta FDI}_{it-j}+ \\ {\sum }_{\mathrm{j}=0}^{q_3}{\Phi }_{{URB}_i}{\Delta URB}_{it-j}+{\sum }_{\mathrm{j}=0}^{q_4}{\Phi }_{{GDP}_i}{\Delta GDP}_{it-j}+{\epsilon }_{it}(\mathsf{\tau }) \end{gathered}$$

(4)

Equation (4) can be expanded using the QARDL-ECM format under the panel QARDL context. This can prevent correlations by projecting ${\epsilon }_{it}$ on ${\mathsf{\Delta }FC}_{it}$, ${\mathsf{\Delta }FDI}_{it}$, ${\mathsf{\Delta }URB}_{it}$, ${\mathsf{\Delta }FC}_{it}$ and ${\mathsf{\Delta }GDP}_{it}$ with the form ${\epsilon }_{it}={\sigma }_{FC}{\Delta FC}_{it}+{\sigma }_{FDI}{\Delta FDI}_{it}+{\sigma }_{URB}{\Delta URB}_{it}+{\sigma }_{GDP}{\Delta GDP}_{it}+{\upsilon }_{it}.$ Consequently, ${\epsilon }_{it}$ is no longer correlated with ${\mathsf{\Delta }FC}_{it}$, ${\Delta FDI}_{it}$, ${\mathsf{\Delta }URB}_{it}$, and ${\mathsf{\Delta }GDP}_{it}$. In the QARDL-ECM model, the long-term link between the dependent and independent variables at the τ-th quantile is represented by the coefficient ($\mathsf{\delta }$) in this model, whilst the short-term dynamics are defined by the coefficients ($\mathsf{\Phi }$). The panel model’s QARDL-ECM version is

$$\begin{gathered} {\mathrm{Q}}_{\mathsf{\Delta }{CO}_{2_{it}}}=\mu \left(\tau \right)+\mathsf{\rho }\left(\mathsf{\tau }\right)({CO}_{2_{it-1}}-{\mathsf{\delta }}_{\mathrm{C}\mathrm{F}}{\left(\tau \right)FC}_{it-1}+{\mathsf{\delta }}_{\mathrm{F}\mathrm{D}\mathrm{I}}{\left(\tau \right)FDI}_{it-1}+{\mathsf{\delta }}_{\mathrm{U}\mathrm{R}\mathrm{B}}{\left(\tau \right)URB}_{it-1}+ \\ {\mathsf{\delta }}_{\mathrm{G}\mathrm{D}\mathrm{P}}{\left(\tau \right)GDP}_{it-1}+{\sum }_{\mathrm{j}=1}^{\mathrm{p}-1}{\Phi }_{{co2}_i}{\left(\tau \right)\Delta co2}_{it-j}+{\sum }_{\mathrm{j}=0}^{q_1-1}{\Phi }_{{FC}_i}{\left(\tau \right)\Delta FC}_{it-j}+ \\ {\sum }_{\mathrm{j}=0}^{q_2-1}{\Phi }_{{FDI}_i}\left(\tau \right){\Delta FDI}_{it-j}+{\sum }_{\mathrm{j}=0}^{q_3-1}{\Phi }_{{URB}_i}\left(\tau \right){\Delta URB}_{it-j}+{\sum }_{\mathrm{j}=0}^{{\mathrm{q}}_4-1}{\Phi }_{{GDP}_i}\left(\tau \right){\Delta GDP}_{it-j}+{\epsilon }_{it}(\mathsf{\tau }) \end{gathered}$$

(5)

The cumulative short-run effect of the preceding CO₂ emissions on the contemporary CO₂ emissions is calculated using the delta method, which is ${{\mathsf{\Phi }}_{{CO}_2}}^{\ast }={\sum }_{\mathrm{i}=1}^{\mathrm{p}-1}\partial {\mathsf{\Phi }}_{{CO}_{2_j}}$. The cumulative short-run impacts of the existing and preceding levels of FC, FDI, URB, and GDP are calculated by ${{\mathsf{\Phi }}_{FC}}^{\ast }={\sum }_{\mathrm{j}=0}^{q_1-1}\partial {\mathsf{\Phi }}_{{FC}_{\mathrm{j}}}$, ${{\mathsf{\Phi }}_{FDI}}^{\ast }={\sum }_{\mathrm{j}=0}^{q_2-1}\partial {\mathsf{\Phi }}_{{FDI}_{\mathrm{j}}}$, ${{\mathsf{\Phi }}_{URB}}^{\ast }={\sum }_{\mathrm{j}=0}^{q_3-1}\partial {\mathsf{\Phi }}_{{URB}_{\mathrm{j}}}$ and ${{\mathsf{\Phi }}_{GDP}}^{\ast }={\sum }_{\mathrm{j}=0}^{{\mathrm{q}}_4-1}\partial {\mathsf{\Phi }}_{{GDP}_{\mathrm{j}}}$, respectively. Similarly, the cointegration among the long-run variables of FC, FDI, URB, and GDP is described as ${{\mathsf{\delta }}_{FC}}^{\ast }=-{\displaystyle \frac{{\delta }_{FC}}{\rho }}, {{\mathsf{\delta }}_{FDI}}^{\ast }=-{\displaystyle \frac{{\delta }_{FDI}}{\rho }}, {{\mathsf{\delta }}_{URB}}^{\ast }=-{\displaystyle \frac{{\delta }_{URB}}{\rho }}$, and ${{\mathsf{\delta }}_{GDP}}^{\ast }=-{\displaystyle \frac{{\delta }_{GDP}}{\rho }}$ correspondingly. It is expected that ρ (the ECM parameter) will be significantly negative. The Wald test assesses the coefficients’ significance to investigate the short- and long-term nonlinear impacts of FC, FDI, URB, and GDP on CO₂.

4. Results and Interpretation

The findings of the various tests and analyses are presented in this section. Before employing the panel QARDL, we first examined the cross-sectional dependency (CD). Selecting appropriate unit root tests that account for CD is essential, as overlooking CD can lead to biased assessments of stationarity and cointegration, size distortions, and unreliable results (Westerlund, 2007). Thus, we employ the Pesaran (2015) CD test to detect cross-section dependency issues. The results are summarized in

Table 2. Cross-sectional dependence and unit root tests.

Variables	CD	CIPS I(0)	CIPS I(1)	Decision
CO2	35.14 ***	−1.978	−5.940 ***	I(1)
FC	14.30 ***	−3.816 ***	−6.128 ***	I(0)
FDI	36.63 ***	−3.310 ***	−4.512 ***	I(0)
URB	49.21 ***	−2.595	−3.203 ***	I(1)
GDP	72.82 ***	−2.414	−4.423 ***	I(1)

Note: ‘***’ denote statistical significance at the 1% significance levels, respectively.

. It shows that all the variables have CD issues, indicating that the associations have accurate calculation difficulty.

To assess the stationarity properties, we used the CIPS unit root test, a second-generation stationarity test. This test is suitable for handling the cross-sectional dependence of the panel variables. All variables are shown in Table 2 to be stationary. Particularly, the unit root test indicates that FC and FDI are stationary at level (I(0)), while CO₂, URB, and GDP are stationary at first difference (I(1)). These results justify the use of the panel QARDL model, which is designed to handle variables with different orders of integration. Once stationarity is confirmed, the subsequent step involves examining the long-run relationship among the variables. Consequently, we conducted a cointegration test, as shown in

Table 3. Cointegration test.

	Statistic-Value	p-Value
Phillips–Perron t (Pedroni test)	−2.5169 ***	0.0059
Augmented Dickey–Fuller t (Pedroni test)	−3.3110 ***	0.0005
Dickey–Fuller t (Kao test)	−1.6816 **	0.0463
Unadjusted modified Dickey–Fuller t (Kao test)	−2.0831 **	0.0186
Unadjusted Dickey–Fuller t (Kao test)	−2.5629 ***	0.0052
Westerlund test	2.7898 ***	0.0026

Note: ‘***’and ‘**’ denote statistical significance at the 1% and 5% significance levels, respectively.

. The findings of Pedroni (1999, 2004), Kao (1999), and Westerlund and Edgerton (2007) confirm the existence of a long-term connection between CO₂ emissions and the explanatory factors. We thus reject the null hypothesis of no cointegration.

Table 4. Panel QARDL estimation results.

Panel A
Quantile	FC	FDI	URB	GDP	$\mathbf{E}\mathbf{C}{\mathbf{T}}_{\mathbf{t}-1}$
10th	0.0016 (0.0707)	0.0245 ** (0.0119)	−0.0479 *** (0.0108)	−2.0964 *** (0.4720)	−0.0046 *** (0.0011)
20th	−0.0021 (0.0058)	0.0219 ** (0.0091)	−0.0448 *** (0.0082)	−1.9975 *** (0.3592)	−0.0044 *** (0.0008)
30th	−0.0039 0.0045	0.0199 *** (0.0070)	−0.0423 *** (0.0064)	−1.9184 *** (0.2791)	−0.0042 *** (0.0007)
40th	−0.0059 * (0.0034)	0.0175 *** (0.0054)	−0.0394 *** (0.0048)	−1.8276 *** (0.2128)	−0.0039 *** (0.0005)
50th	−0.0068 ** (0.0032)	0.0166 *** (0.0051)	−0.0383 *** (0.0046)	−1.7918 *** (0.2002)	−0.0038 *** (0.0005)
60th	−0.0079 ** (0.0032)	0.0153 *** (0.0050)	−0.0368 *** (0.0046)	−1.7424 *** (0.1999)	−0.0038 *** (0.0005)
70th	−0.0089 *** (0.0035)	0.0141 *** (0.0055)	−0.0353 *** (0.0049)	−1.6975 *** (0.2173)	−0.0036 *** (0.0005)
80th	−0.0101 ** (0.0040)	0.0128 ** (0.0063)	−0.0337 *** (0.0057)	−1.6486 *** (0.2507)	−0.0035 *** (0.0006)
90th	−0.0117 ** (0.0051)	0.0108 (0.0080)	−0.0314 *** (0.0073)	−1.5756 *** (0.3181)	−0.0034 *** (0.0007)
Panel B
10th	0.0182 ** (0.0075)	0.0312 ** (0.0129)	−0.0008 (0.0023)	−0.2347 ** (0.0977)
20th	0.0123 ** (0.0058)	0.0266 *** (0.0099)	−0.0009 (0.0018)	−0.2196 *** (0.0749)
30th	0.0068 (0.0044)	0.0223 *** (0.0074)	−0.0010 (0.0013)	−0.2054 *** (0.0558)
40th	0.0017 (0.0034)	0.0184 *** (0.0057)	−0.0011 (0.0010)	−0.1925 *** (0.0431)
50th	−0.0013 (0.0031)	0.0160 *** (0.0052)	−0.0011 (0.0009)	−0.1846 *** (0.0394)
60th	−0.0057 * (0.0032)	0.0125 ** (0.0055)	−0.0012 (0.0010)	−0.1732 *** (0.0419)
70th	−0.0076 ** (0.0035)	0.0111 * (0.0060)	−0.0012 (0.0011)	−0.1685 *** (0.0455)
80th	−0.0095 ** (0.0038)	0.0096 (0.0066)	−0.0012 (0.0012)	−0.1637 *** (0.0502)
90th	−0.0133 *** (0.0047)	0.0066 (0.0082)	−0.0013 (0.0015)	−0.1537 ** (0.0621)

Note: The matrix depicts the outcomes of quantile estimation. Brackets show the standard errors. *, ** and *** show significance at the 10%, 5% and 1%, levels, respectively.

presents the panel QARDL estimation results assessing the impact of fiscal consolidation, foreign direct investment, urbanization, and GDP on CO₂ emissions across OECD countries. Panel A reports the short-run dynamics, while Panel B displays the long-run relationships. To ensure the robustness of the model, diagnostic checks for heteroskedasticity and autocorrelation are summarized in

Table 5. Diagnostic test results.

Test	Statistics [p-Values]
Heteroskedasticity test: Modified Wald test	16.81 [0.4675]
Autocorrelation test: Wooldridge test	0.054 [0.8185]

. The results confirm that the model satisfies key econometric assumptions, supporting the validity and reliability of the estimates. In the short run, the overall effect of FC is negative and statistically significant, except for the 10th–30th quantiles, indicating that fiscal consolidation plays a key role in reducing carbon emissions. However, long-run estimations reveal a varying effect of FC on carbon emissions across quantiles, with statistical significance maintained throughout, excluding quantiles from 0.3 to 0.5. Specifically, at the 0.1 and 0.2 quantiles, FC has a significant and positive impact, with coefficients of 0.0182 and 0.0123. By contrast, at higher quantiles, from 0.6 to 0.9, the effect becomes negative. This suggests that in the long run, when the level of pollution is high in OECD countries, their fiscal consolidation efforts allow them to achieve more effective and sustained reductions in CO₂ emissions, ensuring long-term environmental and economic benefits. This asymmetric long-run effect can be explained by differences in policy mechanisms across pollution levels. At lower quantiles, where emissions are already low, consolidation may reduce public investment in renewable energy and environmental programs, leading to a positive effect on emissions. By contrast, at higher quantiles, where pollution is high, fiscal consolidation is more often accompanied by structural reforms, stricter environmental regulations, and the reallocation of resources toward green initiatives, resulting in significant emission reductions. This suggests that the ecological effectiveness of FC is conditional on the level of emissions. Actually, at low emission levels, fiscal consolidation could affect sustainability; however, it becomes a powerful instrument for achieving both fiscal and environmental goals when pollution pressures are greater. Thus, in the long run, effective fiscal consolidation serves as the key factor for achieving sustained reductions in CO₂ emissions, driving both environmental and economic benefits, particularly in higher quantiles where the negative effect is more notable.

Moreover, effective fiscal consolidation is generally linked to GDP growth. For example, Alesina and Ardagna (2010), Alesina et al. (2002), Giudice et al. (2007), and De Cos and Moral-Benito (2013) show that consolidation episodes tend to improve real GDP growth in developed countries. This growth-enhancing effect provides a channel through which fiscal consolidation may stimulate investment in R&D and foster technological innovation, which in turn supports environmental sustainability.

These results are consistent with Ullah et al. (2021), who claimed that reducing CO₂ emissions has been assisted by technological improvement. These authors contend that R&D and technological innovation can potentially enhance environmental sustainability. Xin et al. (2022) present comparable findings, showing that technological advancements contributed to CO₂ reductions in host economies. Consequently, technological progress fosters the adoption of new technologies that effectively reduce energy consumption. Sun (2022) further elaborates on how technological innovation drives economic optimization and enhances industrial production processes, ultimately leading to a reduction in CO₂ emissions. According to Chen et al. (2023), technological developments create chances for the manufacturing and domestic sectors to employ high-quality products, which tend to lower carbon intensity. The incorporation of a carbon tax in overall fiscal consolidation efforts may also help explain the observed results. Carbon taxes have become an increasingly common tool among OECD countries, as such revenues contribute to deficit reduction while addressing environmental goals (Rausch, 2013).

The GDP coefficient provides a significant and negative effect on CO₂ emissions at all quantiles in the short and long term. This finding also backs up the earlier result, showing that economic development facilitates better governance and technological progress, allowing for environmental protection. Thus, despite continuous industrialization challenges in several OECD countries, economic development is a key factor in enhancing environmental management (Lan et al., 2025). The findings demonstrate that the Environmental Kuznets Curve (EKC) hypothesis is not valid for OECD countries, as there is no inverted U-shaped relationship between economic growth and environmental degradation. The EKC theory states that economic progress first increases energy consumption and environmental deterioration, but beyond a certain threshold of economic growth, it leads to cleaner technologies and more efficient energy use (Sarkodie & Strezov, 2018). In other words, as a country’s income rises, harm to the environment primarily rises, but after reaching a certain level of income, further economic development leads to improvements in environmental quality. The lack of evidence for this relationship suggests that in OECD countries, rising GDP may already reduce CO₂ emissions, likely due to technological progress, cleaner energy use, and stricter environmental policies, making the EKC framework unnecessary.

The effect of urbanization on carbon emissions is significant only in the short run, where it reduces emissions, but becomes insignificant in the long run in OECD countries. There is a study highlighting similarly mixed and context-dependent outcomes regarding the urbanization–emissions nexus. According to Li and Lin (2015), the relationship between urbanization and CO₂ emissions is complex, context-specific and varies across income levels. Urbanization tends to increase emissions in low-, middle- and high-income countries, while it is more likely to reduce emissions in middle- and high-income groups. These heterogeneous findings reflect the role of urbanization and development in shaping environmental outcomes. Consistent with this, our results show that the effect of urbanization is not uniform but differs depending on the level of development and between the short run and the long run. The QARDL results show that in the short run, the effect of urban population on carbon emission is negative and significant at all quantiles. This is consistent with some previous assessments, where urbanization was shown to decrease carbon emissions. For instance, Gnangoin et al. (2023) and Hua et al. (2023) confirm that with the increase in the urban population, people will switch their energy use from traditional fuels to modern energy sources associated with lower carbon emissions (Barnes et al., 2010; Defries & Pandey, 2010; Pachauri & Jiang, 2008). Therefore, energy consumption associated with the expansion of public goods scale effects and the ongoing development of technology can lower carbon emissions. Nevertheless, in the long run, the impact of urbanization on CO₂ emissions in OECD countries becomes insignificant. This result aligns with Ma and Ogata (2024). The study confirms the non-significant effect of urbanization rate on CO₂ emissions in OECD countries by using a System Generalized Method of Moments (SGMM) model to investigate the urbanization rate and carbon dioxide emissions nexus in 136 countries from 1990 to 2020. According to these authors, this result can be attributed to already high levels of urbanization, advanced technology, energy efficiency, and strict environmental policies. The existence of these factors, regardless of urban population growth, makes the relationship between urbanization and emissions less pronounced.

The short and long-run effects of FDI on CO₂ emissions are positive and significant, revealing that increased FDI inflows could result in higher CO₂ emissions. While FDI brings economic benefits, climate change costs might undermine these gains, as FDI often coincides with higher environmental carbon emissions. This view is supported by numerous studies highlighting the tendency of FDI to neglect environmental issues (Zhu et al., 2016; Pao & Tsai, 2011; Cole et al., 2011; Demena & Afesorgbor, 2020).

For the sake of robustness, we re-estimated the model after excluding five OECD countries with comparatively lower GDP per capita (France, Japan, Italy, Spain, and Portugal) based on OECD GDP per capita rankings (OECD, 2022). This helps examine whether the relationship between fiscal consolidation and CO₂ emissions changes depending on a country’s level of economic development and check if the main results are influenced by structural differences in less affluent economies, such as slower growth, higher debt levels, or weaker institutions. The results are reported in

Table A1. Panel QARDL estimation results after excluding lower-GDP countries.

Panel A
Quantile	FC	FDI	URB	GDP	$\mathbf{E}\mathbf{C}{\mathbf{T}}_{\mathbf{t}-1}$
10th	−0.0033 (0.0030)	−0.0001 (0.0005)	0.7022 *** (0.1118)	−0.0031 * (0.0016)	−0.9302 *** (0.0395)
20th	−0.0028 (0.0022)	−0.0001 (0.0004)	0.7117 *** (0.0813)	−0.0029 ** (0.0012)	−0.9364 *** (0.0287)
30th	−0.0030 * (0.0017)	−0.0002 (0.0070)	0.7188 *** (0.0064)	−0.0002 *** (0.0009)	−0.9410 *** (0.0228)
40th	−0.0035 ** (0.0017)	−0.0001 (0.0003)	0.7724 *** (0.0661)	−0.0314 * (0.0187)	−0.9410 *** (0.0224)
50th	−0.0036 ** (0.0016)	−0.0002 (0.0026)	0.7716 *** (0.0637)	−0.0287 (0.0181)	−0.9412 *** (0.0216)
60th	−0.0036 ** (0.0016)	−0.0002 (0.0003)	0.7704 *** (0.0661)	−0.0247 (0.1999)	−0.9412 *** (0.0224)
70th	−0.0037 * (0.0019)	−0.0003 (0.0003)	0.7690 *** (0.0765)	−0.0202 (0.2173)	−0.9412 *** (0.0259)
80th	−0.0037 * (0.0023)	−0.0003 (0.0004)	0.7676 *** (0.0987)	−0.0141 (0.0263)	−0.9412 *** (0.0314)
90th	−0.0037 * (0.0023)	−0.0003 (0.0003)	0.7676 *** (0.0932)	−0.1546 (0.0265)	−0.9413 *** (0.0316)
Panel B
10th	0.0164 ** (0.0066)	0.0029 ** (0.0011)	−0.0053 (0.0043)	−0.0029 *** (0.0008)
20th	0.0109 ** (0.0043)	0.0026 *** (0.0009)	−0.0059 * (0.0034)	−0.0027 *** (0.0006)
30th	0.0069 (0.0044)	0.0025 *** (0.0007)	−0.0065 ** (0.0028)	−0.0026 *** (0.0005)
40th	0.0033 (0.0036)	0.0023 *** (0.0006)	−0.0069 *** (0.0024)	−0.0025 *** (0.0004)
50th	−0.0005 (0.0034)	0.0021 *** (0.0005)	−0.0074 *** (0.0020)	−0.0024 *** (0.0003)
60th	−0.0072 ** (0.0035)	0.0018 *** (0.0006)	−0.0083 *** (0.0021)	−0.0021 *** (0.0004)
70th	−0.0076 ** (0.0035)	0.0017 *** (0.0006)	−0.0085 *** (0.0023)	−0.0021 *** (0.0004)
80th	−0.0121 *** (0.0042)	0.0016 ** (0.0007)	−0.0089 *** (0.0027)	−0.0020 *** (0.0004)
90th	−0.0171 *** (0.0051)	0.0014 (0.0009)	−0.0096 *** (0.0034)	−0.0018 *** (0.0006)

Note: The matrix depicts the outcomes of quantile estimation. Brackets show the standard errors. *, ** and *** show significance at the 10%, 5% and 1%, levels, respectively.

of Appendix A. They indicate that the direction and significance of the estimated coefficients remain consistent in both the short and long run, confirming that the main findings are not driven by less affluent economies. As an additional robustness check, we incorporate an environmental tax as a proxy for eco-taxation, using data obtained from the OECD database. Since this variable is only available from 1994 onward, the robustness analysis covers the period 1994–2020. As shown in

Table A2. Panel QARDL estimation results after including environmental tax.

Panel A
Quantile	FC	FDI	URB	ENVTAX	GDP	$\mathbf{E}\mathbf{C}{\mathbf{T}}_{\mathbf{t}-1}$
10th	−0.1005 (0.0989)	0.4345 *** (0.1369)	−0.1091 *** (0.0317)	−0.5662 (0.5102)	−4.0617 (2.7423)	−8.8386 * (5.4762)
20th	−0.1060 (0.0753)	0.4194 *** (0.1078)	−0.1122 *** (0.0248)	−0.4451 (0.3855)	−5.1448 ** (2.1282)	−8.4377 ** (4.2683)
30th	−0.1224 ** (0.0592)	0.4406 *** (0.0889)	−0.1170 *** (0.0202)	−0.5833 * (0.3063)	−5.6630 *** (1.7278)	−5.6526 * (3.1704)
40th	−0.1262 ** (0.0584)	0.4385 *** (0.0877)	−0.1172 *** (0.0199)	−0.5769 * (0.3023)	−5.7025 *** (1.7056)	−5.7540 * (3.1298)
50th	−0.1473 *** (0.0558)	0.4268 *** (0.0841)	−0.1179 *** (0.0191)	−0.5406 * (0.2897)	−5.9251 *** (1.6341)	−6.3268 ** (2.9970)
60th	−0.1717 *** (0.0572)	0.4132 *** (0.0862)	−0.1188 *** (0.0196)	−0.4987 * (0.2972)	−6.1821 *** (1.6763)	−6.9879 ** (3.0738)
70th	−0.2027 *** (0.0651)	0.3959 *** (0.0981)	−0.1199 *** (0.0223)	−0.4452 (0.3379)	−6.5099 *** (1.9060)	−7.8314 ** (3.4952)
80th	−0.2349 *** (0.0778)	0.3781 *** (0.1177)	−0.1211 *** (0.0268)	−0.3899 (0.4057)	−6.8495 *** (2.2878)	−8.7049 ** (4.1928)
90th	−0.2695 *** (0.0949)	0.3589 ** (0.1438)	−0.1223 *** (0.0328)	−0.3305 (0.4956)	−7.2137 *** (2.7945)	−9.6421 * (5.1198)
Panel B
10th	0.0046 (0.0063)	0.0339 *** (0.0106)	−0.0065 *** (0.0024)	−0.0426 (0.0335)	−0.3039 * (0.1642)
20th	−0.0022 (0.0039)	0.0288 *** (0.0065)	−0.0066 *** (0.0015)	−0.0353 * (0.0205)	−0.3285 *** (0.1002)
30th	−0.0053 * (0.0031)	0.0265 *** (0.0051)	−0.0067 *** (0.0012)	−0.0319 ** (0.0160)	−0.3399 *** (0.0783)
40th	−0.0059 ** (0.0030)	0.0260 *** (0.0049)	−0.0067 *** (0.0011)	−0.0312 ** (0.0154)	−0.3422 *** (0.0752)
50th	−0.0081 *** (0.0027)	0.0244 *** (0.0045)	−0.0068 *** (0.0010)	−0.0288 ** (0.0141)	−0.3503 *** (0.0690)
60th	−0.0100 *** (0.0027)	0.0230 *** (0.0045)	−0.0068 *** (0.0010)	−0.0268 * (0.0144)	−0.3570 *** (0.0705)
70th	−0.0115 *** (0.0029)	0.0219 *** (0.0049)	−0.0068 *** (0.0011)	−0.0252 * (0.0155)	−0.3624 *** (0.0760)
80th	−0.0137 *** (0.0035)	0.0202 *** (0.0058)	−0.0069 *** (0.0013)	−0.0227 (0.0183)	−0.3707 *** (0.0899)
90th	−0.0157 *** (0.0041)	0.0187 *** (0.0068)	−0.0069 *** (0.0016)	−0.0206 (0.0216)	−0.3779 *** (0.1058)

Note: The matrix depicts the outcomes of quantile estimation. Brackets show the standard errors. *, ** and *** show significance at the 10%, 5% and 1%, levels, respectively.

in Appendix A, the findings indicate that the environmental tax exerts a statistically significant effect in both the short and long run, though the strength of this effect varies across quantiles. In the short run, significance appears only at the 10% level and is concentrated between the 30th and 60th quantiles. In the long run, the effect becomes more robust, with significance observed from the 20th to the 70th quantiles. At higher quantiles (80th and 90th), however, the tax effect is not significant, suggesting that when CO₂ emissions reach very high levels, taxation alone is insufficient to counterbalance the structural drivers of pollution.

Fiscal consolidation continues to show negative coefficients in both the short and long run, reinforcing our baseline results. This implies that fiscal instruments, whether through expenditure adjustments or tax measures, can support economic growth while also contributing to the reduction in CO₂ emissions. Taken together, the results highlight the importance of combining fiscal discipline with well-designed environmental taxes. While eco-taxation proves effective at moderate levels of emissions, additional measures, such as green investment, stricter regulations, and incentives for clean technologies, are required to tackle higher levels of pollution that exceed the controlling capacity of taxation alone.

Table 6. Results of the Wald test for the constancy of parameters.

Variables	Wald Statistics [p-Value]
$\mathsf{\rho }$	11.93 *** [0.0006]
${\mathsf{\delta }}_{\mathrm{F}\mathrm{C}}$	4.23 ** [0.0398]
${\mathsf{\delta }}_{\mathrm{F}\mathrm{D}\mathrm{I}}$	3.83 ** [0.0504]
${\mathsf{\delta }}_{\mathrm{U}\mathrm{R}\mathrm{B}}$	1.29 [0.2552]
${\mathsf{\delta }}_{\mathrm{G}\mathrm{D}\mathrm{P}}$	14.63 *** [0.0010]
${\mathsf{\Phi }}_{FC}$	5.31 ** [0.0212]
${\mathsf{\Phi }}_{FDI}$	7.41 *** [0.0065]
${\mathsf{\Phi }}_{URB}$	14.25 *** [0.0002]
${\mathsf{\Phi }}_{GDP}$	23.85 *** [0.000]

Note: ‘***’ and ‘**’ denote statistical significance at the 1% and 5% significance levels, respectively.

displays the results of the Wald test. It examines the dependency between long-run and short-run parameters. Furthermore, the Wald test checks for the nonlinearity of the parameters (Cho et al., 2015). Accepting the null hypothesis implies that the relationship has no asymmetries or nonlinearities. The Wald test confirms that the null hypothesis of symmetry is rejected for all variables in both the long and short run, except for URB in the short run, where the null hypothesis is accepted. This confirms that the estimated parameters exhibit nonlinear and asymmetric relationships. Specifically, the impact of these variables on CO₂ emissions differs significantly across the distribution of CO₂ emissions in OECD countries, validating the relevance of the QARDL framework. Moreover, the rejection of the null for the error correction parameter (ρ) supports the existence of a distribution-dependent long-run equilibrium relationship in the model.

In

Table 7. Granger non-causality test.

Null Hypothesis	$\overline{\mathit{W}}$-Stat.	$\overline{\mathit{Z}}$-Stat.	Prob.	Decisions
FC does not cause CO2	1.9624	2.8060	0.0050	Bidirectional
CO2 does not cause FC	0.4280	−1.6676	0.0954
FDI does not cause CO2	17.5483	6.9592	0.0000	Bidirectional
CO2 does not cause FDI	2.6163	4.7122	0.0000
GDP does not cause CO2	3.7643	8.0593	0.0000	Unidirectional
CO2 does not cause GDP	0.8716	−0.3742	0.7083
URB does not cause CO2	4.9157	11.4163	0.0000	Bidirectional
CO2 does not cause URB	12.1163	32.4093	0.0000

, the Dumitrescu and Hurlin (2012) Granger non-causality test is implemented to evaluate the causal relationships between the variables. This test is appropriate for panel data and considers that all coefficients vary throughout the cross-sections (Ahmed & Shimada, 2019).

$\overline{W}$ statistics and $\overline{Z}$ statistics are the two types of statistics that are considered under this approach. $\overline{Z}$ statistics indicate a conventional normal distribution, whereas $\overline{W}$ statistics averages test statistics. The results reveal bidirectional causality between FC and CO₂, but with notable differences in the strength of relationships. There is weak evidence of reverse causality from CO₂ to FC at the 10% significance level, suggesting a limited feedback effect. This confirms the importance of integrating environmental considerations into fiscal consolidation strategies while also considering the potential effects of environmental shifts on fiscal policy decisions. This finding is consistent with Ike et al. (2020), in which bidirectional causal relationship between fiscal policy and CO₂ emissions was uncovered for Thailand with annual data from 1972 to 2014. FDI and URB both exhibit bidirectional causal relationships with CO₂ emissions. Moreover, a unidirectional causality was found between GDP and CO₂. These results reveal that while foreign direct investment and urban population are mutually linked with carbon emissions, economic development has a more one-sided effect. This requires targeted policy strategies that effectively promote both economic growth and environmental sustainability.

5. Conclusions

The fiscal policy–environmental pollution nexus remains a subject of contention in the wider discussion on globalization and environmental impact. The primary topic of controversy in this debate is whether expansionary or contractionary fiscal policy is more beneficial for the environment. This has contributed to conflicting hypotheses, particularly among those who argue that fiscal contractions lead to decreased production, lowering carbon emissions. The present study considers fiscal consolidation as a potential driver of GDP growth, supported by its positive correlation with economic performance. This is a new and unexamined cause of ecological improvement. To this end, this paper uses a QARDL model to estimate the dynamics of fiscal consolidation shocks on CO₂ emissions through different quantiles in a group of 17 OECD countries from 1987 to 2020. For robustness checks, the Dumitrescu and Hurlin (2012) Granger non-causality test was conducted to examine the causal relationships among the variables.

The findings showed mixed results in both the long run and short run. It shows that fiscal consolidation reduces CO₂ emissions at almost all quantiles in the short run, while in the long run, the effect is positive at lower quantiles and becomes negative at upper extreme quantiles. Additionally, the causal analysis reveals a bidirectional relationship between FC and CO₂, with a stronger influence from FC to CO₂.

Based on these empirical results, it is confirmed that fiscal consolidation programs play a key role in guiding the OECD countries’ economy towards sustainable development. CO₂ emissions reduction is closely associated with incorporating environmental considerations with imperatives for fostering economic growth. Consequently, these programs contribute significantly by directing investments towards sustainable infrastructure and encouraging green technologies. In addition, another explanation for this result lies in the use of carbon tax revenues to support fiscal consolidation efforts. Carbon taxes can enhance emissions reduction initiatives. Hence, channeling carbon tax revenues toward deficit reduction may help explain the beneficial fiscal consolidation effect on carbon emissions, as it aligns environmental goals with fiscal discipline.

This paper reflects the channels through which fiscal policy impacts the environment. It carries useful implications for governments, policymakers, and environmental regulating agencies. While the adoption of a contractionary fiscal policy is generally expected to reduce derived demand and, consequently, lower carbon emission levels, this research reveals a different reality. From a non-Keynesian perspective, implementing fiscal consolidation policies is associated with increased GDP growth, as evidenced by empirical findings for OECD countries. Following this assumption, contrary to the conventional expectation that GDP growth entails higher production and, consequently, increased CO₂ emissions, this study provides evidence indicating that GDP growth driven by fiscal consolidation may contribute to CO₂ emissions reduction. While the effect tends to be beneficial overall, the long-run estimation results reveal a more nuanced picture that needs further investigation. The robustness analysis further confirms our main findings. By excluding the lower-GDP countries from the sample, the results remained consistent across quantiles in terms of sign and statistical significance, which shows that the relationship between fiscal consolidation and CO₂ emissions is not driven by these countries.

These findings confirm the nuanced long-run relationship, where fiscal consolidation significantly increases emissions at lower quantiles and reduces them at higher quantiles. This reinforces the importance of aligning fiscal strategies with long-term environmental sustainability goals, particularly in low-emission contexts where fiscal consolidation may otherwise lead to increased emissions. In this context, strengthening consolidation strategies to further reduce CO₂ emissions becomes a key factor. Addressing fiscal consolidation challenges through managing public spending and increasing carbon tax revenue can play a vital role in promoting growth and supporting climate policies. This can potentially support the transition to cleaner energy and more responsible use of fossil fuels.

The effect of fiscal consolidation on CO₂ emissions in OECD nations is the main topic of this study. Although this analysis offers insightful information, it ignores the possible impact of other policy aspects that could also affect emissions, such as monetary policy or energy transition strategies. Furthermore, because fiscal consolidation episodes identified through the narrative approach were not yet accessible after 2020, this research is limited to that period. These limitations open promising areas for future research, particularly by considering the connections between fiscal consolidation and other policy instruments and extending the analysis to a more recent period once updated data becomes available.