Evaluation of the Relationship Between Ecological Footprint, Economic and Political Stability Variables in SAARC Countries with PVAR Analysis

Abstract

South Asia faces the dual challenge of sustaining rapid economic growth while managing severe ecological pressures. This study explores the relationship between Ecological Footprint (EF), Financial Development (FD), Economic Growth (GDP), Foreign Direct Investment (FDI), and Political Stability (PS) in SAARC countries from 2000 to 2020. Using a Panel Vector Autoregression (PVAR) combined with a Vector Error Correction Model (VECM), the analysis captures both short-run dynamics and long-run equilibrium relationships, addressing endogeneity among variables. Results reveal that EF negatively correlates with FD, GDP, and FDI, while showing a positive association with PS. Cointegration tests using dynamic and fully modified ordinary least squares confirm long-term relationships between the variables. Impulse response functions illustrate how shocks to one variable affect others over time, highlighting complex interactions. Granger causality tests suggest limited short-term causal links, reflecting the multifaceted nature of these relationships. This research is particularly relevant as SAARC countries face the dual challenge of sustaining rapid economic growth while mitigating ecological pressures. The study advances the literature by explicitly integrating political stability into the environmental–economic nexus, a factor often overlooked in earlier regional analyses. By providing empirical evidence on the joint role of economic, financial, and political drivers of ecological sustainability, the paper contributes both to academic debate and to the design of more balanced policy frameworks for sustainable development in South Asia.

1. Introduction

Both developed and developing countries face significant challenges in balancing economic growth with protecting the global environment. In recent years, the relationship between economic development and environmental sustainability has become a focal point for policymakers and researchers alike [1]. As the world struggles with the dual challenge of promoting economic growth while reducing environmental degradation, it is important to understand the relationship between various economic indicators and ecological footprints.

The ecological footprint (EF) is a widely used indicator that measures the environmental pressure exerted by human activities [2]. Originally conceptualized in the 1990s, EF accounts for the amount of land and water required to sustain consumption and absorb waste. Unlike traditional measures such as CO₂ emissions, EF provides a more comprehensive view of environmental stress [3,4,5,6,7]. Previous studies have reported mixed findings: while some argue that economic and financial development improve environmental quality through technological progress and regulatory frameworks, others suggest that these factors exacerbate ecological stress, highlighting the need for further empirical investigation [8]. Economic growth (proxied by GDP), financial development (credit to the private sector) [9], and foreign direct investment (FDI) are typically regarded as engines of growth, but they may also carry hidden environmental costs. In addition, political stability—often assumed to facilitate economic progress—can have complex implications for ecological sustainability. For example, rigid political systems may discourage environmental reform [10], while instability can undermine long-term investments in sustainable infrastructure.

Against this backdrop, the main objective of this study is to evaluate the interrelationships between ecological footprint, economic performance, and political stability in South Asian Association for Regional Cooperation (SAARC) countries during the period 2000–2022. Specifically, the study aims to:

• Assess the dynamic links between ecological footprint, economic growth, financial development, FDI, and political stability.
• Identify both short-run and long-run effects among these variables using advanced panel econometric techniques.
• Provide policy-relevant insights for achieving a balance between economic growth and ecological sustainability in the SAARC region.

The contribution of this study is threefold. First, it enriches the existing literature by incorporating both economic and political dimensions into the analysis of ecological footprint, an approach less common in prior studies. Second, it applies a Panel Vector Autoregression (PVAR) framework, which allows for the exploration of dynamic feedback effects without imposing strict causality assumptions [11]. Third, by focusing on the SAARC region, the study sheds light on a group of countries facing common challenges of rapid economic growth, resource constraints, and political volatility, thereby offering region-specific insights.

The scientific novelty of this research lies in explicitly integrating political stability as a key determinant of ecological sustainability, addressing a gap in prior SAARC-focused studies. By jointly examining the dynamic interactions of economic growth, financial development, FDI, and political stability on ecological outcomes, the study provides new empirical evidence on the interplay between political, economic, and environmental factors. These findings not only contribute to the academic literature but also offer practical guidance for designing balanced policies aimed at sustainable development in the region.

This study employs data from the World Development Indicators (WDI) and the Global Footprint Network (GFN) for the period 2000–2022. The methodology includes panel unit root tests, cointegration tests [12,13], the Vector Error Correction Model (VECM), Impulse-Response Functions, and Granger causality tests to examine both short-term and long-term interactions. In addition, robustness is checked through Fully Modified OLS (FMOLS) and Dynamic OLS (DOLS) estimations [14]. The strengths of the PVAR model lie in its ability to capture static and dynamic interdependencies, address cross-sectional connections, and account for heterogeneity across countries while mitigating the limitations of short time-series panels. These characteristics make PVAR particularly suitable for analyzing the complex nexus of political, economic, and environmental variables in the SAARC context. While the findings highlight a complex and asymmetric relationship between economic indicators and ecological sustainability, they also suggest that political stability may play a mitigating role in reducing environmental pressure. Nevertheless, as with any empirical study, limitations remain. Data constraints, particularly in terms of availability and consistency across countries, and the reliance on proxy indicators for financial development and political stability, should be considered when interpreting the results. Despite these limitations, the study provides important insights for designing integrated policies that balance economic objectives with ecological sustainability in the SAARC region.

2. Literature Review

Numerous studies have examined the complex relationships between economic variables and ecological sustainability. For instance, ref. [1] studied the relationship between energy use, per capita income, and ecological footprint in MINT countries (Mexico, Indonesia, Nigeria, and Türkiye) between 1976 and 2016. Their study utilized a vector error correction model and panel vector autoregression analysis to show that there is a positive correlation between energy use and ecological footprint. This finding underlines the environmental challenges caused by increasing energy consumption in developing economies and highlights the urgent need for policies that promote sustainable energy practices and reduce ecological degradation. Similarly, ref. [15] investigated the impact of real income, financial development, and trade openness on the ecological footprint, finding that these factors positively influence the ecological footprint in major emitting countries. Furthermore, ref. [3] used Ecological Footprint analysis to study High, Middle, and Low-Income countries from 1961 to 2005, revealing that high-income countries increased per capita Footprint due to resource consumption, while low-income countries reduced per capita Footprint despite population growth, indicating more efficient resource use. These results highlight contrasting trends in resource consumption and environmental impact among income groups, underscoring the need for sustainable development strategies.

Building on this ref. [16], examined political instability in MENA and its impact on the HDI-EF relationship. Using panel VAR and data from 16 countries (1990–2016), they find an upside-U relationship between HDI and EF, noting that political instability, especially post-2011 uprisings, affects this relationship. This emphasizes the need for stable governance and effective policies to promote human development as well as environmental sustainability. Additionally, ref. [17] found that positive shocks to political stability significantly increase the ecological footprint in both the short and long run, while negative shocks have minimal impact, revealing ecological asymmetry. Another related study by ref. [18] in South Asia found that corruption and political instability significantly affect carbon and ecological footprints. Corruption has a long-term positive impact on both footprints but a short-term negative effect on the carbon footprint. Political instability positively influences both footprints in the long term but shows a negative and insignificant relationship with the carbon footprint in the short term. The study emphasizes that addressing corruption and enhancing political stability is essential for improving both carbon and ecological footprints. Furthermore, ref. [19] comparative study on South Asian countries’ ecological footprints highlights significant disparities. India has the largest footprint at 1063.37 million global hectares (gha), with Sri Lanka the smallest at 24.14 million gha. Nepal shows the highest per-person footprint at 3.56 gha, contrasting with Bangladesh’s lowest at 0.62 gha. All countries exceed their biocapacity, with India leading in ecological overshoot (−469.04 million gha), followed by Nepal and Pakistan. emphasizing the urgent need for targeted environmental policies and sustainable development initiatives to address escalating ecological challenges and promote long-term sustainability.

Several studies have also analyzed the complex relationship between economic growth and environmental sustainability in different countries and contexts—for example, ref. [10] they studied the relationship between energy consumption, economic growth, and ecological footprint in countries dependent on energy imports from Russia. Using econometric techniques like the Panel Vector Autoregression (PVAR) model, they found that an energy shock from Russia could negatively impact economic growth while encouraging a shift towards renewable energy, positively affecting the ecological footprint. Similarly, ref. [20] conducted a comprehensive analysis in Nigeria, revealing a positive association between GDPs per capita and economic growth, indicating that as economic activity expands, so does the ecological footprint. Extending this inquiry to Tunisia [21], demonstrated a unidirectional causality between economic growth, and CO₂ emissions and a bi-directional link between CO₂ emissions and GDP, suggesting that economic growth contributes to environmental degradation through increased emissions. Furthermore, ref. [4] focused on Qatar, finding a U-shaped relationship between ecological footprint and real GDP per capita, supporting the Environmental Kuznets Curve hypothesis that initial economic growth may degrade the environment before potentially improving beyond a certain income threshold. Extending this analysis to 15 MENA countries, it highlighted varied impacts of economic development and socio-political factors on ecological footprint, noting that energy consumption exacerbates ecological footprint differently across oil-exporting and non-oil-exporting nations. Further exploring Nigeria, ref. [20] examined interactions between economic growth, energy consumption, population dynamics, and ecological footprint, identifying complex interdependencies. Similarly, ref. [22] analyzed financial development, economic growth, and non-renewable energy consumption across top emitter countries, showing these factors negatively impact environmental quality by increasing ecological footprint. Ref. [23], investigated Turkey, finding a U-shaped causality from energy consumption to ecological footprint, suggesting initial increases followed by potential decreases due to shifts towards energy-intensive industries—finally, ref. [24] studied 120 countries, revealing that renewable energy consumption positively influences economic growth and ecological quality, whereas non-renewable energy consumption promotes economic development but increases the ecological footprint [20,25].

Research on Foreign Direct Investment (FDI) underscores its dual impact on ecological footprints, influenced by technological advancements and environmental regulations. For instance, ref. [26] revealed significant bidirectional relationships between FDI and Ecological Footprint across 44 member countries from 1990 to 2016, using Panel Vector Autoregression (PVAR), stressing the necessity of managing FDI to foster sustainable economic growth while curbing environmental degradation. Similarly, ref. [20] highlighted a positive relationship between FDI and ecological footprints across 92 countries from 2001 to 2016, signaling potential environmental risks with higher FDI levels. Ref. [27], noted disparities in how FDI affects ecological outcomes across income groups and sectors, with High-Income countries benefiting as ecological havens while Middle-Income countries bear more production-related burdens. Moreover, ref. [28] emphasized integrating FDI policies with energy conservation efforts, demonstrating in Jiangsu Province how FDI impacts energy consumption intensity through scale, structure, and technology effects. Additionally, ref. [29] found a positive association between industrialization and ecological footprints in the G-7 economies from 1991 to 2018, suggesting that nuclear energy production, industrialization, fossil fuel energy consumption, and FDI contribute to increased ecological footprints in these countries. These studies collectively highlight the complex interplay between FDI, economic activities, and environmental sustainability, underscoring the need for targeted policies to manage ecological footprints globally [30].

Turning to the impact of financial development on ecological footprints, recent studies have explored the intricate relationship between economic factors and environmental sustainability across various global contexts. Ref. [22], explored the influence of financial development and economic growth on the ecological footprint in the top 10 emitter countries, finding that financial development, economic growth, and non-renewable energy consumption contribute negatively to environmental quality by increasing the ecological footprint. In a related study, ref. [31] examined technological innovation, financial development, and economic growth in West Asia and Middle East (WAME) countries, revealing that while technological innovation decreases the ecological footprint, financial development tends to increase it, alongside urbanization which exacerbates ecological degradation in the region. Similarly, ref. [32] focused on Belt and Road Initiative (BRI) countries, showing significant positive relationships between financial development, economic growth, energy consumption, FDI, and ecological footprint, underscoring financial activities’ role in environmental degradation. Moreover, ref. [33] globally analyzed financial development’s impact on ecological footprint, finding a negative association that suggests improved environmental quality with increased financial development, contingent upon other factors like energy consumption and GDP. Additionally, ref. [34] explored the multidimensional impact of financial development on ecological footprints globally, highlighting an inverted-U-shaped relationship were initial negative effects transition to positive outcomes, urging environmentally conscious financial practices. Finally, ref. [35] investigated Malaysia, revealing short-term negative impacts of financial sector development on ecological footprint, moderated by institutional quality to mitigate long-term environmental degradation, emphasizing the dual importance of financial sector development and institutional quality in environmental policy formulation and implementation.

3. Methodology and Data

3.1. Data Description

This study utilizes annual panel data covering eight member countries of the South Asian Association for Regional Cooperation (SAARC)—Afghanistan, Bangladesh, Bhutan, India, Maldives, Nepal, Pakistan, and Sri Lanka—over the period 2000–2022. The SAARC region was chosen for three main reasons. First, these countries represent one of the world’s fastest-growing yet resource-constrained regions, where rapid economic growth has coincided with increasing environmental pressures. Second, the region is highly vulnerable to ecological imbalances due to high population density, dependence on natural resources, and political volatility. Third, SAARC countries have committed to regional cooperation on sustainability and development, making them an important case for analyzing the interplay between ecological sustainability, economic dynamics, and governance.

The analysis investigates the impact of selected economic, financial, and political variables on ecological sustainability. All data are obtained from two reputable sources: the World Development Indicators [36], provided by the World Bank, and the Global Footprint Network [37]. The selected time frame allows for the identification of long-term trends and structural relationships in the region. The dependent variable in this study is the Load Capacity Factor (LCF), which serves as a proxy for ecological sustainability. LCF is defined as the ratio of a country’s biocapacity to its ecological footprint, both measured in global hectares. A value greater than one indicates ecological reserve, while a value below one implies ecological deficit. Data for this indicator is retrieved from the Global Footprint Network and is widely used in the literature as a comprehensive measure of environmental pressure and sustainability performance. To assess the drivers of ecological sustainability, several key independent variables are incorporated into the analysis. Economic Growth (GDP) is measured using the annual growth rate of real Gross Domestic Product, expressed in constant 2010 US dollars. This variable captures the level of economic expansion in each country and is sourced from the [36]. Financial Development (FD) is represented by the domestic credit to the private sector as a percentage of GDP. This indicator reflects the depth and maturity of the financial system and its capacity to allocate resources efficiently. Foreign Direct Investment (FDI) is included as an indicator of global economic integration and is measured by net FDI inflows, expressed as a percentage of GDP. FDI is often linked to both positive and negative environmental impacts, depending on the nature of investment and the regulatory framework of the host country. Finally, Political Stability and Absence of Violence (PS) is used to capture the quality of governance and institutional stability. These variable measures the perceived likelihood of political unrest, violence, or terrorism, which may significantly affect both environmental policies and investment flows. The data for PS are drawn from the Worldwide Governance Indicators and accessed via the [36], database.

Altogether, these variables provide a multidimensional framework for analyzing how economic performance, financial activity, foreign investment, and political conditions collectively influence ecological sustainability in SAARC countries.

Table 1. Variables description.

Variable	Definition	Unit	Source
(LNLCF)	Load Capacity Factor	Total Ecological Footprint (in global hectares)	[38]
FD	Financial Development	Domestic credit to private sector by banks (% of GDP)	[37]
GDP	Economic growth	GDP Growth Annual	[37]
FDI	Foreign direct investment	Foreign direct investment, net inflows (% of GDP)	[37]
PS	Political Stability	Political Stability No Violence	[37]

in the following section presents a detailed overview of the variables, including their definitions, measurement units, data sources, and summary statistics.

In Figure 1 below, the dataset of countries examined in this study is presented. It is apparent from the graphs that Nepal shows significant differences compared to other countries. The data from Nepal exhibit a more volatile pattern, indicating fluctuations that are distinct from those observed in other countries. This volatility may reflect the country’s relatively unstable economic and political environment, which has influenced both financial and environmental indicators. In contrast, most of the other SAARC countries display comparatively smoother trends, highlighting the unique trajectory of Nepal within the region.

Table 2. Descriptive statistics and correlation.

	LNLCF	GDP	FD	FDI	PS
Mean	16.54250	5.454467	33.53699	1.974353	0.265395
median	16.28624	5.765822	32.32837	0.923246	0.246048
Maximum	20.01403	37.68719	103.5262	16.78347	0.453906
Minimum	13.20212	−32.90883	2.435703	−0.638806	0.192247
Std. Dev.	1.890520	5.665245	18.46310	2.971381	0.055831
Skewness	−0.052520	−0.735541	0.656002	2.725771	1.318490
Kurtosis	2.530339	21.06648	4.059208	10.49088	4.779766
Jarque–Bera	1.659906	2354.693	20.37682	615.1336	72.53548
Probability	0.436070	0.000000	0.000038	0.000000	0.000000
Sum	2845.309	938.1683	5768.362	339.5886	45.64795
Sum Sq. Dev.	611.1653	5488.245	58291.54	1509.777	0.533030
Observations	172	172	172	172	172
Correlation
	LNLCF	GDP	FD	FDI	PS
LNLCF	1.000000
GDP	−0.050236	1.000000
FD	−0.179982 *	−0.080895	1.000000
FDI	−0.569696 ***	0.102874	0.006884	1.000000
PS	0.072526	0.038971	−0.319270 ***	−0.093127	1.000000

Notes: *** denotes significance at the 1% level and * at the 10% level.

presents descriptive statistics and correlation relationships among the variables. The upper section of the table displays descriptive statistics for each variable. It is observed that the FD variable exhibits a notably high standard deviation, indicating significant variability in its values. The lower section of the table examines correlation coefficients between the variables. The LNLCF series shows a negative correlation with the GDP, FD, and FDI variables, suggesting that as GDP, FD, and FDI values increase, LNLCF tends to decrease. Additionally, there is a positive correlation between LNLCF and PS. The GDP variable correlates negatively with FD and positively with FDI and PS, indicating that higher GDP values are associated with higher FDI and PS values. These findings contribute to a better understanding of the relationships and variations among the variables in the dataset.

3.2. Model Specification

To analyze the main factors affecting environmental quality in SAARC countries, this study uses a Panel Vector Autoregression (PVAR) model. Unlike traditional regression approaches that require a strict distinction between dependent and independent variables, the PVAR model treats all variables as potentially endogenous and mutually interdependent. This is particularly appropriate in the current context, where ecological, economic, financial, and political factors interact dynamically and are influenced by both domestic policies and external shocks. Alternative approaches, such as static panel regressions, dynamic ARDL models, or system GMM estimators, have been widely used in the literature. However, these methods often impose restrictive assumptions regarding causality, exogeneity, or homogeneity across units. For example, ARDL frameworks are well-suited for small-sample time series but cannot adequately capture cross-country interdependencies, while system GMM is designed for large panels and may suffer from instrument proliferation in shorter panels. By contrast, PVAR provides a balance between flexibility and robustness, allowing for feedback effects, country-specific heterogeneity, and the dynamic propagation of shocks.

Another key advantage of PVAR is its ability to account for both internal (endogenous) and external (exogenous) shocks, a feature of particular importance for small open economies like those in the SAARC region. This flexibility enables robust estimation whether analyzing one country individually or a group of countries jointly. In this study, PVAR is employed to explore the interactions among ecological footprint, economic growth, financial development, foreign direct investment, and political stability.

Formally, the PVAR model is specified as follows:

$$\begin{array}{c}{y_i}_t=y_{it-1}{A}_1+y_{it-2}{A}_2+\dots .+{y_{it-p}}_{+1}{A}_{p-1}+y_{it-p}{A}_p+x_{it}B+u_{it}\\ +e_{it};iϵ\left(\mathrm{1,2},N\right),tϵ\left(\mathrm{1,2}\dots T\right)\end{array}$$

(1)

where Y_it is the (1 × k) vector of dependent variables, X_it is the (1 × k) vector of exogenous variables, and u_it and e_it are (1 × k) dimensional, constant effects, and idiosyncratic error terms vectors. With A₁, A₂, …, A_p−1, A_p, shown in the equation, the (1 × k) dimensional B matrix is estimated.

The PVECM model, as an extension of the PVAR model, allows for simultaneous analysis of long-run equilibrium relationships and short-run dynamics among variables. This model captures the short-term adjustment process of variables in response to deviations from their long-run equilibrium, providing a more comprehensive and realistic analysis of panel data.

The model equations for our variables are as follows:

$$\begin{array}{c}\mathsf{\Delta }{{\mathrm{lnLCF}}_i}_t=c_{1i}+{\sum }_{j=1}^q{{\beta }_{11}}_{\mathit{ij}}\mathsf{\Delta }{{\mathrm{lnLCF}}_{\mathit{it}}}_{-j}+{\sum }_{j=1}^q{\beta }_{12\mathit{ij}}\mathsf{\Delta }{\mathrm{FD}}_{\mathit{it}-j}+{\sum }_{j=1}^q{\beta }_{13\mathit{ij}}\mathsf{\Delta }{\mathrm{GDP}}_{\mathit{it}-j}+{\sum }_{j=1}^q{\beta }_{14\mathit{ij}}\mathsf{\Delta }{\mathrm{FDI}}_{\mathit{it}-j}+\\ {\sum }_{j=1}^q{\beta }_{15\mathit{ij}}\mathsf{\Delta }{\mathrm{PS}}_{\mathit{it}-j}+{\beta }_{16i}{{\epsilon }_i}_{t-1}+u_{1\mathit{it}}\end{array}$$

(2)

$$\begin{array}{c}\mathsf{\Delta }{{\mathrm{FD}}_i}_t=c_{2i}+{\sum }_{j=1}^q{{\beta }_{21}}_{\mathit{ij}}\mathsf{\Delta }{{\mathrm{lnLCF}}_{\mathit{it}}}_{-j}+{\sum }_{j=1}^q{\beta }_{22\mathit{ij}}\mathsf{\Delta }{\mathrm{FD}}_{\mathit{it}-j}+{\sum }_{j=1}^q{\beta }_{23\mathit{ij}}\mathsf{\Delta }{\mathrm{GDP}}_{\mathit{it}-j}+{\sum }_{j=1}^q{\beta }_{24\mathit{ij}}\mathsf{\Delta }{\mathrm{FDI}}_{\mathit{it}-j}+{\sum }_{j=1}^q{\beta }_{25}\mathit{ij}\mathsf{\Delta }{\mathrm{PS}}_{\mathit{it}-\mathit{j}}+\\ {\beta }_{26i}{{\epsilon }_i}_{t-1}+u_{2\mathit{it}}\end{array}$$

(3)

$$\begin{array}{c}\mathsf{\Delta }{{\mathrm{GDP}}_i}_t=c_{3i}+{\sum }_{j=1}^q{{\beta }_{31}}_{\mathit{ij}}\mathsf{\Delta }{{\mathrm{lnLCF}}_{\mathit{it}}}_{-j}+{\sum }_{j=1}^q{\beta }_{32\mathit{ij}}\mathsf{\Delta }{\mathrm{FD}}_{\mathit{it}-j}+{\sum }_{j=1}^q{\beta }_{33\mathit{ij}}\mathsf{\Delta }{\mathrm{GDP}}_{\mathit{it}-j}+{\sum }_{j=1}^q{\beta }_{34\mathit{ij}}\mathsf{\Delta }{\mathrm{FDI}}_{\mathit{it}-j}+\\ {\sum }_{j=1}^q{\beta }_{35}\mathit{ij}\mathsf{\Delta }{\mathrm{PS}}_{\mathit{it}-j}+{\beta }_{36i}{{\epsilon }_i}_{t-1}+u_{3\mathit{it}}\end{array}$$

(4)

$$\begin{array}{c}\mathsf{\Delta }{{\mathrm{FDI}}_i}_t=c_{4i}+{\sum }_{j=1}^q{{\beta }_{41}}_{\mathit{ij}}\mathsf{\Delta }{{\mathrm{lnLCF}}_{\mathit{it}}}_{-j}+{\sum }_{j=1}^q{\beta }_{42\mathit{ij}}\mathsf{\Delta }{\mathrm{FD}}_{\mathit{it}-j}+{\sum }_{j=1}^q{\beta }_{43\mathit{ij}}\mathsf{\Delta }{\mathrm{GDP}}_{\mathit{it}-j}+{\sum }_{j=1}^q{\beta }_{44\mathit{ij}}\mathsf{\Delta }{\mathrm{FDI}}_{\mathit{it}-j}+{\sum }_{j=1}^q{\beta }_{45}\mathit{ij}\mathsf{\Delta }{\mathrm{PS}}_{\mathit{it}-j}+\\ {\beta }_{46i}{{\epsilon }_i}_{t-1}+u_{4\mathit{it}}\end{array}$$

(5)

$$\begin{array}{c}\mathsf{\Delta }{{\mathrm{PS}}_i}_t=c_{5i}+{\sum }_{j=1}^q{{\beta }_{51}}_{\mathit{ij}}\mathsf{\Delta }{{\mathrm{lnLCF}}_{\mathit{it}}}_{-j}+{\sum }_{j=1}^q{\beta }_{52\mathit{ij}}\mathsf{\Delta }{\mathrm{FD}}_{\mathit{it}-j}+{\sum }_{j=1}^q{\beta }_{53\mathit{ij}}\mathsf{\Delta }{\mathrm{GDP}}_{\mathit{it}-j}+{\sum }_{j=1}^q{\beta }_{54\mathit{ij}}\mathsf{\Delta }{\mathrm{FDI}}_{\mathit{it}-j}+{\sum }_{j=1}^q{\beta }_{55}\mathit{ij}\mathsf{\Delta }{\mathrm{PS}}_{\mathit{it}-j}+\\ {\beta }_{56i}{{\epsilon }_i}_{t-1}+u_{5\mathit{it}}\end{array}$$

(6)

where the lag value is q, the first difference operator is Δ, the error correction term is ε, and the random error term is u. The VECM is calculated using the unrelated regression (SUR) method, which supports cross-sectional correlations in the residuals and cross-sectional-specific coefficient vectors. The overall econometric procedure applied in this study is illustrated in Figure 2.

3.3. Econometric Procedures

In order to comprehensively examine the dynamic interrelationships between ecological footprint (LNLCF), gross domestic product (GDP), financial development (FD), foreign direct investment (FDI), and political stability (PS) across SAARC countries over the period 2000–2022, a systematic econometric strategy is adopted. This approach follows a sequential logic that ensures methodological rigor while addressing the common challenges associated with macro-panel data, such as cross-sectional dependence, non-stationarity, cointegration, and endogeneity.

To begin with, it is essential to determine whether cross-sectional dependence exists among the panel units. In panel datasets involving multiple countries, the presence of common shocks, regional spillovers, or interconnected economic structures often induces cross-sectional dependence. If unaccounted for, this dependence can bias standard test statistics and lead to misleading inferences. Therefore, the procedure starts with the application of Pesaran’s CD test, which provides a robust diagnostic for detecting cross-sectional correlations across units [38]. Given the evidence of cross-sectional dependence, the analysis proceeds with second-generation panel unit root tests that are specifically designed to accommodate such interdependencies. Unlike first-generation tests—such as Levin, Lin, and Chu [39], or Im, Pesaran, and Shin [40], which assume cross-sectional independence, second-generation tests offer more reliable results in the presence of common factors. Accordingly, as in ref. [41], the Cross-sectionally Augmented IPS (CIPS) test and [42] PANIC approach are employed to determine the order of integration of the series. Establishing the stationarity properties of the variables is a crucial prerequisite for assessing the existence of long-run equilibrium relationships.

Following confirmation that the variables are integrated of the same order, the analysis turns to testing for cointegration. The existence of a cointegrating relationship implies that although the individual variables may be non-stationary, their linear combination is stable over time, reflecting a long-run equilibrium. To test cointegration, two complementary approaches are applied: Pedroni’s [12], residual-based cointegration tests and the Johansen Fisher Panel cointegration test [43,44]. These methodologies provide robust and flexible tools for examining long-run relationships in heterogeneous panels. Once cointegration is established, the appropriate modeling framework becomes the Vector Error Correction Model (VECM), which can capture both short-run dynamics and long-run equilibrium adjustments. The VECM framework, as introduced by Engle and Granger [45], integrates the short-term fluctuations with the long-run steady-state path through an error correction mechanism. Before estimating the model, the optimal lag length must be selected to ensure dynamic consistency. This selection is based on information criteria such as the Akaike Information Criterion (AIC), Schwarz Bayesian Criterion (SBC), Final Prediction Error (FPE) and Hannan-Quinn information criterion (HQ) [46,47], which help balance model complexity and goodness of fit.

To further investigate the dynamic interactions among the variables over time, Impulse Response Functions (IRFs) are derived from the estimated VECM. IRFs trace the time path of the dependent variables in response to one-unit shocks in each of the explanatory variables, thereby providing valuable insights into the magnitude, direction, and persistence of inter-variable effects [47]. This enables a nuanced understanding of how the system responds to exogenous changes.

In parallel, short-run causal linkages among the variables are assessed using Granger causality tests within the VECM framework. These tests evaluate whether past values of one variable contain information that helps predict another, thereby uncovering potential causal mechanisms [45]. The results of these tests complement the impulse response analysis and contribute to a more comprehensive understanding of the short-term interdependencies. Finally, to obtain robust long-run parameter estimates, the Fully Modified Ordinary Least Squares (FMOLS) and Dynamic Ordinary Least Squares (DOLS) estimators are employed. Both estimators are designed to address potential issues of serial correlation and endogeneity in cointegrated systems. FMOLS modifies the standard least squares approach by incorporating corrections for endogeneity and autocorrelation [48], whereas DOLS includes leads and lags of the differenced regressors to produce unbiased and efficient estimators [49]. The combination of these two methods ensures the credibility and robustness of the long-run coefficient estimates.

In sum, the econometric procedure employed in this study follows a well-structured sequence of diagnostic testing and model estimation, ensuring that both short-run and long-run dynamics are properly captured. This rigorous approach provides a solid empirical foundation for analyzing the complex interactions among environmental, economic, and institutional variables in the context of SAARC countries.

4. Results

Given the strong evidence of cross-sectional dependence obtained from both the Pesaran CD and Breusch-Pagan LM tests, the study refrains from interpreting first-generation unit root test results, as their assumptions are violated in the presence of such dependence. Instead, second-generation methods—specifically, PANIC and CIPS—are employed to examine the stationarity characteristics of the panel data.

The PANIC test decomposes the panel into common and idiosyncratic components, providing a detailed view of non-stationarity dynamics. As summarized in

Table 3. Summary of PANIC Test Results.

Variable	Number of Non-Stationary Common Factors	Stationary Idiosyncratic Components (Count)
LNLCF	0 out of 6	7 out of 8 (AFG, BGD, BTN, IND, MDV, NPL, PAK)
GDP	0 out of 6	1 out of 8 (LKA)
FD	0 out of 6	4 out of 8 (AFG, LKA, NPL, PAK)
FDI	0 out of 6	3 out of 8 (BTN, PAK, BGD)
PS	0 out of 6	0 out of 8

, all variables—GDP, Financial Development (FD), Foreign Direct Investment (FDI), Political Stability (PS), and the log of Ecological Footprint (LNLCF), exhibit six non-stationary common factors, with none identified as stationary, suggesting strong shared trends across countries. The p-values for the common component tests are consistently high (close to 1), confirming the presence of unit roots in the common factors. However, the idiosyncratic elements show mixed stationarity across countries and variables. Specifically, LNLCF shows significant stationarity in most countries, with several cross-sections rejecting the unit root null hypothesis, while GDP largely fails to reject it except for Sri Lanka, indicating persistence of unit roots. Financial Development rejects the unit root hypothesis in Afghanistan, Sri Lanka, Nepal, and Pakistan, while FDI shows partial stationarity with Bangladesh, Bhutan, and Pakistan rejecting the null. The pooled idiosyncratic tests provide evidence against the null of no cointegration for LNLCF, FD, and FDI, indicating some degree of stationarity in their idiosyncratic components, whereas GDP and PS fail to reject this null.

Complementing these findings, the Pesaran CIPS test (

Table 4. Summary of Pesaran CIPS Test Results.

Variable	CIPS Stat.	Trunc. CIPS Stat.	p-Value	Stationary Cross Sections (Count)	Order of Integration
LNLCF	−3.13510	−2.95260	<0.01	4 of 8 (AFG **, BGD *, BTN ***, PAK **)	I(0)
GDP	−3.63354	−3.63354	<0.01	5 of 8 (AFG ***, BTN ***, MDV ***, NPL *, PAK *)	I(0)
FD	−2.14274	−2.14274	≥0.10	1 of 8 (PAK ***)	I(1)
FDI	−2.20411	−2.20411	≥0.10	2 of 8 (IND ***, LKA **)	I(1)
PS	−2.55984	−2.55984	<0.05	2 of 8 (AFG **, IND ***)	I(0)

Note: Superscripts denote CADF significance: * p < 0.10, ** p < 0.05, *** p < 0.10.

) provides a panel unit root framework that also accounts for cross-sectional dependence. The results align closely with the PANIC test. LNLCF and GDP are stationary at levels in most cross-sections, while PS is stationary in only two of the eight countries. On the other hand, FD and FDI remain non-stationary, confirming the need for first differencing in these variables. These results further reinforce the robustness of the PANIC test findings.

Together, the PANIC and CIPS results present a coherent narrative: while non-stationarity is pervasive in the common components, several idiosyncratic components are stationary across different countries. These findings justify the subsequent application of first differencing and cointegration techniques to capture both short-term dynamics and long-run relationships in the model. Accordingly, the presence of cointegration among the variables is further examined using the Johansen Fisher Panel Cointegration Test. As reported in

Table 5. Johansen Fisher Panel Cointegration Test.

Trend assumption No deterministic trend (restricted constant)
Series: LNLCF GDP FD FDI PS
Lags interval (in first differences): 1 to 3
Unrestricted Cointegration Rank Test (Trace)
Hypothesized		Trace	0.05
No. of CE(s)	Eigenvalue	Statistic	Critical Value	Prob. **
None *	0.307568	102.9756	76.97277	0.0002
At most 1	0.191783	51.51922	54.07904	0.0830
At most 2	0.113079	21.70982	35.19275	0.6150
At most 3	0.031899	4.909922	20.26184	0.9834
At most 4	0.002649	0.371325	9.164546	0.9986
Trace test indicates 1 cointegrating eqn(s) at the 0.05 level
Unrestricted Cointegration Rank Test (Maximum Eigenvalue)
Hypothesized		Max-Eigen	0.05
No. of CE(s)	Eigenvalue	Statistic	Critical Value	Prob. **
None *	0.307568	51.45642	34.80587	0.0002
At most 1 *	0.191783	29.80940	28.58808	0.0347
At most 2	0.113079	16.79990	22.29962	0.2452
At most 3	0.031899	4.538597	15.89210	0.9246
At most 4	0.002649	0.371325	9.164546	0.9986

The max-eigenvalue test indicates two cointegrating eqn(s) at the 0.05 level. * Denotes rejection of the hypothesis at the 0.05 level and ** MacKinnon-Haug-Micheli.s (1999) p-values.

, the results indicate the existence of stable long-term relationships among the log of ecological footprint (LNLCF), GDP, FD, FDI, and PS. The trace test identifies at least one cointegrating vector, with the trace statistic (102.9756) exceeding the 5% critical value (76.97277) and yielding a p-value of 0.0002. In parallel, the maximum eigenvalue test rejects the null of no cointegration at the 5% level for two cointegrating vectors, with corresponding statistics of 51.45642 and 29.80940 and p-values of 0.0002 and 0.0347, respectively. These findings suggest that although the individual series are non-stationary in levels, they share a common stochastic trend in the long run. This justifies the use of the Vector Error Correction Model (VECM) to simultaneously capture the short-run dynamics and long-run equilibrium relationships among the variables, with the inclusion of an error correction term that reflects the speed of adjustment toward equilibrium [50].

4.1. Estimation Vector Error Correction Model (VECM)

Based on the results of the cointegration tests, it is evident that the series exhibits a unit root and maintains a long-term relationship among the variables. This necessitates the application of the Vector Error Correction Model (VECM) to estimate the cointegrating vectors accurately. Before employing VECM, it is essential to determine the optimal lag length, which can significantly impact the model’s accuracy. In

Table 6. VAR Lag Order Selection Criteria.

Endogenous Variables: LNLCF GDP FD FDI PS/Exogenous Variables: C
Lag	LogL	LR	FPE	AIC	SC	HQ
0	−1047.120	NA	199.2009	19.48370	19.60787	19.53404
1	−318.2637	1376.728	0.000435	6.449327	7.194364 *	6.751412 *
2	−284.4060	60.81848	0.000371	6.285295	7.651196	6.839118
3	−255.8195	48.70292	0.000349 *	6.218879 *	8.205643	7.024439
4	−241.2703	23.44032	0.000430	6.412413	9.020040	7.469711
5	−224.2093	25.90748	0.000510	6.559431	9.787922	7.868467
6	−191.0194	47.32628 *	0.000453	6.407767	10.25712	7.968541
7	−165.7692	33.66690	0.000474	6.403134	10.87335	8.215645

* Indicates lag order selected by the criterion. LR: sequential modified LR test statistic (each test at 5% level), FPE: Final prediction error, AIC: Akaike information criterion, SC: Schwarz information criterion, HQ: Hannan-Quinn information criterion.

, various criteria such as the Final Prediction Error (FPE) and the Akaike Information Criterion (AIC) suggest that a lag length of three is appropriate. This selection is crucial as it guides the estimation process, ensuring a robust analysis for the Panel Vector Autoregression (PVAR).

Table 7. Vector Error Correction Estimates (α and β Vectors).

π = αβ
Error Correction:	D(LNLCF)	D(GDP)	D(FD)	D(FDI)	D(PS)
$\alpha$	−0.002073	0.065600	−0.179259	−0.008706	0.001690
	(0.00087)	(0.11383)	(0.08754)	(0.02694)	(0.00026)
	[−2.37343]	[0.57628]	[−2.04763]	[−0.32317]	[6.42283]
$\beta$	LNLCF(-1)	GDP(-1)	FD(-1)	FDI(-1)	PS(-1)	C
	1.000000	−0.200115	−0.028137	0.589957	−104.3887	5.254052
		(0.29471)	(0.03941)	(0.23902)	(18.0553)	(4.95141)
		[−0.67903]	[−0.71401]	[2.46822]	[−5.78160]	[1.06112]

Standard errors in ( ) and t-statistics in [ ].

displays the Vector Error Correction Model (VECM) estimates for three lags, indicating the presence of at least one cointegrated vector among the variables LNLCF (Ecological Footprint), GDP, FD, FDI and PS (Political Stability). According to [51], and [52], the π matrix, where β represents the cointegration parameters and α represents the adjustment coefficients, is used to evaluate long-run relationships. Significant coefficients in the π matrix confirm these relationships at the 5% significance level. The Error Correction Term (α Vectors) shows that LNLCF, FD, and PS are statistically significant. Specifically, LNLCF and FD have negative significant coefficients, indicating that they adjust negatively to return to long-run equilibrium, while PS has a positive significant coefficient, suggesting that it adjusts positively.

$$\mathrm{LNLCF}=0.200115\mathrm{GDP}+0.028137\mathrm{FD}-0.589957\mathrm{FDI}+104.3887\mathrm{PS}-5.254052$$

(7)

In the Cointegration Equation (β Vectors), with LNLCF as the dependent variable, it is realized that FDI is positively significant, indicating that increased foreign direct investment is associated with a higher ecological footprint, while PS is negatively significant, suggesting that greater political stability is linked to a lower ecological footprint. Conversely, GDP and FD are not significant in this context. This analysis highlights the robust long-term interdependencies among LNLCF, FDI, and PS.

Table 8. Short-Term Estimates of LNLCF as a dependent variable.

	D(LNLCF)	D(GDP)	D(FD)	D(FDI)	D(PS)
D(LNLCF(-1))	−0.471250 ***	−15.17049	−5.265820	−0.104109	0.007083
D(LNLCF(-2))	−0.129647	−13.12593	−5.646720	−1.782139	0.015032
D(LNLCF(-3))	−0.016290	4.998629	−3.799353	−2.334345	0.009594
D(GDP(-1))	−0.000809	−0.890244 ***	0.111054	0.007122	0.000574 **
D(GDP(-2))	−0.000900	−0.782363 ***	0.143930	−0.002511	0.000251
D(GDP(-3))	−0.000289	−0.595013 ***	0.134966	0.017622	0.000181
D(FD(-1))	−0.000116	0.394516 ***	0.217002 **	0.017262	0.000932 ***
D(FD(-2))	−0.001003	−0.042623	0.005751	−0.025146	−0.000117
D(FD(-3))	0.000326	−0.200212	−0.033983	0.071792 **	0.000484
D(FDI(-1))	−0.000726	−1.706025 ***	0.393866	−0.572409 ***	−0.001396
D(FDI(-2))	−0.003299	−0.499913	−0.312977	−0.265278 **	−2.57E-05
D(FDI(-3))	−0.001724	0.547471	−0.349817	0.113384	0.000464
D(PS(-1))	−0.002979	−27.31637	5.219199	−4.188386	−0.108792
D(PS(-2))	0.238553	32.23595	−10.42676	−4.734858	0.008924
D(PS(-3))	−0.119944	−43.21846	−13.44217	2.512090	−0.171508 ***
R-squared	0.215334	0.689601	0.181059	0.346710	0.279265
Adj. R-squared	0.120415	0.652053	0.081993	0.267683	0.192080
Sum sq. resids	0.226022	3837.781	2269.808	214.9209	0.020503
S.E. equation	0.042694	5.563259	4.278423	1.316523	0.012859
F-statistic	2.268605	18.36571	1.827667	4.387226	3.203113
Log likelihood	251.3620	−430.4219	−393.6580	−228.6553	419.3655
Akaike AIC	−3.362315	6.377456	5.852257	3.495076	−5.762364
Schwarz SC	−3.026127	6.713643	6.188444	3.831264	−5.426176
Mean dependent	0.008571	−0.086112	1.235993	0.035515	−0.004137
S.D. dependent	0.045522	9.431320	4.465405	1.538434	0.014306
Determinant resid covariance (dof adj.)		0.000226
Determinant resid covariance		0.000123
Log likelihood		−363.0293
Akaike information criterion		6.414704
Schwarz criterion		8.221713

Note: *, **, and *** indicate significance at the 5%, 1%, and 0.1% levels, respectively.

presents the short-term estimates derived from the Vector Error Correction Model (VECM), with LNLCF as the dependent variable. The coefficients represent the immediate impact of changes in GDP, FD, FDI, and PS on LNLCF over a lag of up to three periods. Significant coefficients, indicated by t-statistics greater than two, highlight the short-term dynamics and interdependencies among these variables. Among the 12 statistically significant coefficients, 4 show positive effects at the 1%, and 10% levels, highlighting their impact on LNLCF. Conversely, 8 coefficients show statistically significant negative effects at the 5% and 10% levels. For instance, the coefficient for D(FDI) (−0.104109) indicates the effect of the previous period’s Foreign Direct Investment (FDI) on the current period’s LNLCF. The t-statistic of [−0.03734] reflects no statistical significance. In contrast, the D(PS) (0.000574) coefficient illustrates the effect of Political Stability from three periods ago on the current period’s LNLCF. The t-statistic of [2.59568] suggests statistical significance at the 1% level, indicating that PS from three periods ago has a notable impact on the current period’s LNLCF. The remaining coefficients follow a similar trend and contribute to a deeper understanding of the relationships between these economic factors.

4.2. Impulse-Response Functions

The impulse response functions in Figure 3 provide insights into the dynamic relationships between ecological footprint (LNLCF), GDP, financial development (FD), foreign direct investment (FDI), and political stability (PS) over time. For LNLCF, its initial negative response to its own shocks stabilizes over time. LNLCF’s response to GDP is characterized by fluctuations, while it initially reacts negatively to FD, eventually turning positive. LNLCF consistently exhibits a negative response to FDI, and to PS, it shows an extremely positive reaction initially, followed by a mildly positive one. Regarding GDP’s responses, it fluctuates in response to LNLCF, initially responds negatively to its own shocks before fluctuating, and shows initial positive reactions to FD followed by fluctuations. Both GDP’s reactions to FDI and PS start negative before displaying fluctuations. FD’s response to LNLCF is initially negative, turning mildly positive, while its response to GDP is marked by fluctuations. FD remains stable in response to its own shocks, and its initial positive reaction to FDI turns negative. FD consistently shows a positive response to PS. In terms of FDI, its response to LNLCF is initially negative, followed by fluctuations, while its responses to GDP and FD start negatively before stabilizing. FDI’s response to its own shocks begins negatively and then stabilizes, and it shows initial negative fluctuations followed by positive fluctuations in response to PS. Lastly, PS initially responds positively to LNLCF, becoming stable, and eventually negative, while its response to GDP is marked by fluctuations. PS’s response to FD is initially positive, turning negative, and it responds positively to FDI. PS’s own shocks are extremely negative.

4.3. Short-Run and Long-Run Dynamics: Granger Causality

This study implements the VEC Granger Causality/Block Exogeneity Wald test [45], to identify short-run causal relationships among key economic variables. The vector autoregression lag size is set at two, as determined by the Schwarz information criterion [46]. The null hypothesis in the causality test states that “X does not Granger-cause Y” [53].

Table 9. VEC Granger Causality/Block Exogeneity Wald Tests.

Excluded	Chi-sq	df	Prob.
Dependent variable: D(LNLCF)
D(GDP)	1.408247	3	0.7036
D(FD)	0.870146	3	0.8326
D(FDI)	0.997618	3	0.8018
D(PS)	1.577883	3	0.6644
All	5.579570	12	0.9358
Dependent variable: D(GDP)
D(LNLCF)	2.808825	3	0.4220
D(FD)	10.41835	3	0.0153
D(FDI)	27.59253	3	0.0000
D(PS)	4.972763	3	0.1738
All	42.64069	12	0.0000
Dependent variable: D(FDI)
D(LNLCF)	0.854620	3	0.8364
D(GDP)	0.853996	3	0.8365
D(FD)	4.237720	3	0.2369
D(PS)	1.050333	3	0.7891
All	6.547280	12	0.8860
Dependent variable: D(PS)
D(LNLCF)	0.299860	3	0.9601
D(GDP)	7.459059	3	0.0586
D(FD)	12.15535	3	0.0069
D(FDI)	3.742518	3	0.2906
All	16.23813	12	0.1806

presents the results of the tests, which assess whether the lagged values of one variable can predict changes in another, indicating causality. For the dependent variable D(LNLCF), none of the variable’s GDP, FD, FDI, or PS show Granger causality. Similarly, LNLCF is not Granger-cause GDP. However, FD and FDI significantly affect GDP at the 5% and 1% levels, respectively, highlighting the role of financial development and foreign direct investment in driving short-term GDP changes. For D(FDI), no significant causality is detected from any variable. For the dependent variable D(PS), financial development (FD) shows a significant Granger-causal relationship at the 1% level, indicating its influence on political stability in the short run. Additionally, GDP has a marginal effect on PS at the 10% significance level. In summary, the Granger causality tests emphasize the short-run significance of financial development and foreign direct investment in influencing GDP and political stability, suggesting that these factors are critical drivers within the model’s context.

4.4. Long-Run Dynamics: FMOLS and DOLS Estimators

Examining long-run dynamics through panel fully modified OLS (FMOLS) and dynamic OLS (DOLS) estimators, this study explores the impact of economic variables on ecological footprints using canonical cointegrating regression (CCR) principles [54,55]. Employing asymptotic unbiased and normal distributions for strong estimation, DOLS allows for exceptions with static OLS, while FMOLS integrates Johansen’s maximum likelihood strategy [51,56] to capture system dynamics, as detailed in

Table 10. DOLS and FMOLS Estimators (LNLCF).

Dependent Variable: LNLCF
Method: Panel Dynamic Least Squares (DOLS)
Panel method: Pooled estimation
Cointegrating equation deterministic: C
Fixed leads and lags specification (lead = 1, lag = 1)
Coefficient covariance computed using default method
Long-run variance (Bartlett kernel, Newey–West fixed bandwidth) used for coefficient covariances
Variable	Coefficient	Std. Error	t-Statistic	Prob.
GDP	−0.002574	0.004166	−0.617749	0.5402
FD	−0.000311	0.000888	−0.349599	0.7285
FDI	−0.004630	0.005356	−0.864435	0.3925
PS	−0.752362	0.224726	−3.347913	0.0018
R-squared	0.999893	Mean dependent var		16.54148
Adjusted R-squared	0.999606	S.D. dependent var		1.894116
S.E. of regression	0.037615	Sum squared resid		0.056595
Long-run variance	0.000390
Dependent Variable: LNLCF
Method: Panel Fully Modified Least Squares (FMOLS)
Panel method: Pooled estimation
Cointegrating equation deterministics: C
Coefficient covariance computed using default method
Long-run covariance estimates (Bartlett kernel, Newey–West fixed bandwidth)
Variable	Coefficient	Std. Error	t-Statistic	Prob.
GDP	0.001185	0.001184	1.001032	0.3184
FD	0.001057	0.000638	1.655989	0.0998
FDI	−0.015855	0.004369	−3.629029	0.0004
PS	−0.831284	0.176052	−4.721809	0.0000
R-squared	0.999014	Mean dependent var		16.54458
Adjusted R-squared	0.998942	S.D. dependent var		1.892871
S.E. of regression	0.061557	Sum squared resid		0.575962
Long-run variance	0.006695

with LCF as the dependent variable.

In Table 10, both DOLS and FMOLS estimations for LNLCF show notable insights. Within the DOLS model, Political Stability (PS) emerges as statistically significant at the 1% level (Prb = 0.0018), contrasting with GDP and Financial Development (FD) which lack statistical significance. PS’s negative coefficient suggests a detrimental effect on LCF, although its elasticity coefficient implies a minor impact. Conversely, FMOLS displays positive coefficients for GDP and FD, albeit nonsignificant, while Foreign Direct Investment (FDI) and PS exhibit significantly negative coefficients at the 1% level. These results indicate adverse impacts on LCF (ecological footprint) from FDI and PS. PS’s substantial coefficient underscores its pivotal role in reducing LCF, highlighting its critical importance in environmental sustainability efforts. This comparative analysis provides nuanced insights into how different econometric methodologies elucidate ecological footprint dynamics, offering valuable implications for policy formulation and sustainable development strategies [14].

5. Discussion

This study investigates the relationship between the ecological footprint and selected economic and political variables in SAARC countries over the 2000–2020 period. Employing advanced panel econometric techniques—including the Vector Error Correction Model (VECM), Granger causality tests, and long-run estimators such as Fully Modified Ordinary Least Squares (FMOLS) and Dynamic Ordinary Least Squares (DOLS)—the analysis examines both short-run dynamics and long-run equilibria among key variables: economic growth, financial development, foreign direct investment (FDI), and political stability.

The empirical findings suggest that economic growth is positively associated with the ecological footprint in the long run. The coefficient (0.200115) from the VECM error correction term implies that rising income levels exert environmental pressure, likely driven by increased production and consumption activities. However, the Granger causality test shows not statistically significant short-run impact of GDP on the ecological footprint (p = 0.7036), and both FMOLS and DOLS estimators show insignificant long-run coefficients. These results lend partial support to the Environmental Kuznets Curve (EKC) hypothesis, which posits that environmental degradation initially worsens with economic growth before improving at higher income levels. Empirical studies provide mixed evidence: while [15,21] found positive GDP-environment relationships in developing contexts, ref. [4] reported a U-shaped EKC in Qatar, and [22] observed that financial development and economic growth jointly increase ecological pressure. These variations underline the complexity and nonlinearity of the growth-environment nexus. Financial development is found to exacerbate ecological degradation in the long term. The positive coefficient (0.028137) in Equation (7) suggests that an expanding financial sector may indirectly promote environmentally harmful activities. Granger causality tests show no significant short-run relationship (p = 0.8326), while FMOLS results indicate a weakly significant positive long-run effect. DOLS’ findings, however, show a negative but statistically insignificant coefficient. These mixed outcomes align with [22,31] who documented context-dependent effects of financial development on environmental quality, particularly in emerging economies. The ambiguous nature of these results emphasizes the importance of regulating financial flows to avoid environmentally adverse investments.

In contrast, foreign direct investment (FDI) appears to reduce the ecological footprint in the long run. The VECM coefficient (−0.589957) suggests that sustained FDI inflows may introduce cleaner technologies or enhance production efficiency. While the short-run Granger causality test reports no significant relationship (p = 0.8018), FMOLS estimates reveal a significant negative coefficient (−0.015855, p = 0.0004), indicating a beneficial long-term impact. These findings echo those of [26], who noted bidirectional effects of FDI and ecological footprint, and highlight the role of policy in steering FDI toward environmentally responsible sectors. Conversely, refs. [20,29] cautioned against the environmental risks of unregulated FDI in developing countries. The heterogeneity in results—depending on country income level and policy frameworks—calls for region-specific strategies that align investment with sustainability goals.

Political stability exerts the most pronounced effect on ecological outcomes. The positive coefficient (104.3887) in the VECM suggests that political instability is associated with increased environmental degradation, likely due to weakened institutional capacity and ineffective environmental governance. Although no short-run causality is detected (p = 0.6644), long-run estimations from both DOLS (−0.752362, p = 0.0018) and FMOLS (−0.831284, p < 0.0001) indicate a strong and significant negative relationship between political stability and ecological footprint. These findings are supported by [16], who linked post-Arab Spring instability with environmental deterioration, and [18], who demonstrated that political instability raises ecological and carbon footprints in South Asia. Such evidence reinforces the critical role of governance in environmental stewardship. Taken together, the results highlight the diverse and asymmetric impacts of economic and political variables on the ecological footprint in SAARC countries. While economic growth and financial development tend to increase environmental stress over time, their short-run effects appear limited. FDI, in contrast, emerges as a potential vehicle for environmental improvement, particularly in the long run. Most significantly, political stability is shown to be a key factor in fostering sustainable environmental outcomes. These findings underscore the need for a comprehensive policy approach that balances economic development with environmental protection. Strengthening financial regulation, incentivizing green FDI, and promoting political stability are essential components for achieving sustainability in the region.

6. Conclusions and Policy Recommendations

The findings of this study show a complex and asymmetric relationship between ecological sustainability and its economic and political determinants in the SAARC region. Economic growth, financial development, and foreign direct investment (FDI)—while essential for regional progress—are found to increase ecological degradation, underscoring the environmental costs of growth-driven development strategies. In contrast, political stability acts as a mitigating factor, strengthening institutional capacity and contributing to improved environmental outcomes. The confirmed cointegration among variables demonstrates that these interactions are structurally embedded in SAARC economies. Furthermore, impulse-response analysis indicates that ecological footprints react more strongly to shocks in economic variables than to political ones, while the limited short-run causal links emphasize the need for long-term, systemic interventions rather than short-term adjustments.

Overall, the study illustrates that the SAARC region stands at a strategic crossroad. While striving for economic advancement, the region must also confront the growing environmental costs of its development paradigm. The results call for a paradigm shift—one that envisions growth not merely as an increase in economic output but as progress aligned with environmental integrity and institutional quality. Against this backdrop, several policy directions are warranted to reshape the region’s trajectory toward sustainability. First and foremost, environmental considerations must be systematically integrated into macroeconomic frameworks. This means embedding ecological metrics within national accounting systems and ensuring that economic policies are calibrated to prevent further degradation of natural resources. Economic incentives should be designed to reward sustainability, and green investments should be promoted through favorable regulatory and fiscal instruments. Governments and financial institutions must support green entrepreneurship, eco-innovation, and low-carbon technologies. At the same time, the importance of institutional capacity and political stability cannot be overstated. Stable and accountable governance structures should enforce environmental regulations, improve transparency, and facilitate participatory decision-making. Environmental education, public awareness campaigns, and inclusive dialogue are equally vital to cultivating a societal culture that values sustainability.

Regional collaboration must be strengthened, as many ecological challenges transcend national borders. SAARC countries should invest in coordinated environmental governance, including joint monitoring systems, data sharing, and harmonized regulatory approaches. Moreover, investing in environmental research, clean energy infrastructure, and the development of green technologies will help the region transition from resource-intensive growth models to environmentally resilient economies.

Finally, fiscal and trade policies should be restructured to align economic activities with ecological goals. Tools such as carbon pricing, removal of environmentally harmful subsidies, and the establishment of ecological taxes can guide market behavior in a sustainable direction. This study does not merely diagnose a problem; it offers a path forward. The evidence presented here should serve as both a wake-up call and a guidepost for policymakers, scholars, and civil society in the SAARC region. Environmental sustainability cannot remain a peripheral concern—it must be mainstreamed into the heart of economic planning and governance. A sustainable future for SAARC is achievable, but it demands bold, coordinated, and forward-looking action. If prosperity is to be meaningful and enduring, it must be built upon the foundations of ecological balance and institutional resilience.