1 Introduction

From the beginning of the world, a lot has evolved. Climate is the first and most significant factor. Sustainability and ecological stewardship are topics that are becoming more and more prominent by the day. This is accurate, given that many economies throughout the globe are working to prevent atmospheric damage and maintain sustainable progress (Abid et al., 2023a; Qayyum et al., 2021; Yang et al., 2021). Numerous research have been done so far to investigate this problem, but the results are still not particularly compelling. The complexity of the problem and the direct or indirect effects of several external variables have made combating global warming extremely difficult (Abid et al., 2023b; Ali & Kirikkaleli, 2022; Shabir et al., 2022). As authorities in many countries find themselves unable to address the issue while being aware of the crisis's cause and resolution, the gravity of this worry is growing worse. In addition to greenhouse gas (GHG) emissions caused by direct consumption, which are increasing (Yang et al., 2020), unrestrained economic activity in both developed and developing countries is also increasing GHG emissions, particularly carbon dioxide (CO2) emissions (Ali et al., 2022a, 2022b; Fan et al., 2020; Yang et al., 2022; Yuan et al., 2022).

Ecological challenges are developing despite a growing body of information on how human activity is causing environmental devastation and climate change. To stop atmospheric degradation, it becomes more important than ever to look outside the box and take other factors like education and awareness into account. Previous research indicates a relationship between education, ecological consciousness, and pro-environment actions, suggesting that human capital centered on education and return on education may affect ecological sustainability. For example, Chankrajang and Muttarak (2017) contend that people's behaviour toward the adoption of renewable energy goods is influenced by their human capital. Furthermore, there is no denying the need for education to help people comprehend the root factors of worldwide climate change and its dire ramifications (UNESCO, 2010). Zen et al. (2014) provide evidence of how education has a good impact on recycling efforts. Godoy et al. (1998) contend that education lessens forest loss, but Desha et al. (2015) find that education impacts people's choices in adhering to ecological legislation. According to Bano et al. (2018), energy efficiency is one way that human capital contributes significantly to the reduction of CO2 emissions. As reported by Ahmed et al. (2020), increasing human capital can have a big impact on cutting pollution.

In contrast, natural resources have a dual impact on the climate. First, natural resources are utilized for production and consumption; nevertheless, the atmosphere of the nation is impacted by the irresponsible use of natural resources, including extraction, cultivation, and deforestation. Waste and toxins are released into the air and water during the exploitation of natural resources (Hassan et al., 2019). Employing sustainable production and consumption practices decreases natural resource loss and allows resource redevelopment (Ulucak & Ozcan, 2020). On the other hand, economic expansion from natural resources affects environmental quality. Growing economic expansion accelerates natural resource exploitation, increasing pollution (Panayotou, 1993). Natural resources drive gross domestic product (GDP) growth in many nations (Hailu & Kipgen, 2017). Economic progress promotes industrialization, natural resource use, ecological shortfall, and waste (Sarkodie & Strezov, 2018). Forest loss, water scarcity, and climate change enhance natural resource exploitation and generate environmental issues in developing and industrialized nations (Baloch et al., 2019).

In light of this, academics are now concentrating on sustainable natural resources and human capital. Therefore, the focus of the current study is on how human capital and natural resources affect environmental quality in South Asian nations. With about 25% of the world's population living in South Asia, it is one of the most populated areas in the globe. The socioeconomic well-being of the communities of Bangladesh, Bhutan, India, Maldives, Nepal, Pakistan, and Sri Lanka is negatively impacted by a range of energy and ecological issues (Pandey & Asif, 2022; World Bank, 2019). The human capital accumulation in South Asia has exhibited a consistent upward trend, notwithstanding differences in the performance of particular countries. The amount and quality of human capital in these nations have significantly improved, enabling them to grow economically through higher productivity and greater per capita income (Islam, 2020). Besides, the region of South Asia is well known for having an abundance of total natural resources, which include fossil fuels and the possibility of renewable energy sources. This reasoning holds for the nations of South Asia, as natural resources have a non-linear tendency yet grow over time. The use of these resources has a substantial impact on the energy composition of these economies, which in turn has an impact on the emission profiles of each of them (Wang et al., 2024).

The study makes a number of contributions. For instance, to our knowledge, no prior research has looked at how human capital and natural resources affect the environment within the framework of South Asian nations. There is disagreement in the literature on human capital and natural resources which makes it evident that there are advantages and disadvantages to human capital and natural resources, and it is impossible to assess the overall impact without doing a thorough analysis (Ahmed & Wang, 2019; Ahmed et al., 2020; Danish & Hassan, 2023; Haini, 2021; Safdar et al., 2022). As a result, this study makes several contributions to the body of current knowledge. First, we examine how the environment, human capital, and natural resources relate to the nations under study. Second, we employ ecological footprint (EF) as a metric for ecological effect, in contrast to most other research. This is because EF is seen to be a complete indication of environmental harm as it reveals the impact of human activity on the air, water, and soil (Ahmed Memon et al., 2022; Arif et al., 2023; Gill et al., 2023; Qayyum et al., 2024; Rehman et al., 2023). Figure 1 shows how the ecological footprint is rising yearly in South Asia. Third, by elucidating the interaction effect—that is, the combined influence of natural resources and human capital—on EF across South Asian countries, this research adds to the body of existing work. The relationship between human capital, natural resources, and environmental pollution in different regions has been the subject of several studies (Ahmed et al., 2020; Wang et al., 2024). Nevertheless, no study has examined the combined influence of natural resources and human capital on EF across South Asian countries. This analysis is important since the South Asia region, which is home to almost 25% of the world's population, is a key producer, consumer, and exporter of coal, natural gas, oil, and other vital natural resources. In this situation, human capital in the manner of a trained and knowledgeable labor force may, via responsible and technological mining, alter and reconfigure the pattern of resource exploitation and usage in a more eco-friendly way. Finally, we have used some of the more powerful econometrics approaches (second-generation methods), which produce reliable findings even when heterogeneity, serial correlation, and cross-sectional dependency are present.

Fig. 1
figure 1

Source: Global Footprint Network

Ecological Footprint.

Full size image

Following is how the rest of our investigation unfolds. The "literature review" is the second portion. The "model, data, & methodology" is the third. "Results and discussion" is the fourth subsection. "Conclusions and policy implications" make up the fifth part.

2 Literature review

During past and recent periods, various empirical econometrics studies examined the link between natural capital, human capital, and ecological poverty. This research is carrying out different studies about the association of natural resources, human capital, and environmental pollution. The study revealed affirmatively contrary relationships among natural resources, human capital, and ecological quality in different economies.

2.1 Human capital and environmental quality

Many academics use various econometric techniques to examine time series and panel data to evaluate the relationships between human capital and environmental quality. Whether human capital typically raises pollution levels is an open question, as these studies yield conflicting conclusions. Du et al. (2022) investigated the impression of monetary presence and human capital on ecological superiority by analyzing unindustrialized economies. The research used panel data from 2004 to 2019 using advanced econometrics techniques, proving that human capital in these nations has a confident association with environmental quality. Ganda (2022) checked the environmental impacts of human capital in the BRICS by using data from 1990 to 2017, which proved that human capital is a positive affirmation of environmental quality. Human capital and environmental quality both devise bidirectional effects on each other. The study proposed that the government should implement the study that promotes the human skills that lead to green evolution in economies.

Chen et al. (2022) explored the correlation between human capital, urbanization, and ecological footprint and utilized panel data from one hundred and ten economies from 1990 to 2016. Conclusions display that primarily, from a worldwide viewpoint; human capital upsurges first and then reduces the environmental impact. Moreover, Ahmed et al. (2020) checked the linkage between suburbanization, human capital, and EF in G7 economies. The study cast off the progressive level of panel data from 1971 to 2014. Studies indicate that urbanization, GDP per capita, imports, and energy have a positive relationship with ecological footprint, while human capital negatively affects ecological footprint. Rafique et al. (2022) explored the top ten economically complex countries with links among economic complexity, human capital, renewable energy production, economic development, export quality, trade, and environmental impact. This study fully implemented fully modified ordinary least squares, dynamic ordinary least squares, and system generalized method of moments estimators from 1980–2017. Long-term forecasts show that economic difficulty, economic evolution, export superiority, trade, and increased urbanization have positive environmental effects, and human capital and renewable energy production have negative impacts on pollution. Twum et al. (2021) examined environmental efficiency values in various Asia–Pacific areas between 1990 and 2018, which were calculated using a super-efficiency data envelopment analysis model with acceptable inputs and an undesired output. The environmental efficiency indices for South East Asia, South Asia, and East Asia are compared in the study. Findings concluded that human capital encourages environmental efficiency in the main panel and the rest of the regions. Nathaniel (2021) examined the influence of bio-capacity, human capital, and urbanization on ecological footprint. The outcomes suggest that biocapacity, economic progress, and urbanization increase EF while human capital decreases EF. Mahmood et al. (2019) discovered the association between human capital and CO2 emissions using time series data from 1980 to 2014. Results exposed that human capital diminishes CO2 releases. In contrast, using panel estimators, Haini (2021) investigated the relationship between human capital and carbon emissions in the ASEAN economies from 1996 to 2019. The findings indicate that since human capital development influences growth indirectly, it causes a rise in emissions levels.

2.2 Natural resources and environmental quality

Mixed findings are found in research that examines the connections between natural resources and environment degradation using panel and time-series data and various econometric techniques. For instance, Hassan et al. (2019) used the ARDL model on the yearly data from 1970 to 2014 and investigated the effect of economic development and natural resources on Pakistan's EF. The study revealed that natural resources have an affirmative association with an ecological footprint, which damages ecological excellence. Nathaniel et al. (2021) investigated the connection among natural resources, globalization, urbanization, and ecological deprivation in Latin America and Caribbean countries from 1990 to 2017 by using the advanced panel method. The results confirm that the variables mentioned above add to CO2 emissions. Zafar et al. (2019) looked at how natural resources affected the environment in the USA from 1970 to 2015. They found that the USA's environmental quality is enhanced by its natural resources. To study the relationship between natural resources and CO2 emissions for the BRICS, Baloch et al. (2019) utilized the augmented mean group (AMG) approach to data covering the years 1990–2015. They found that natural resources worsen South Africa's environment, but not in Russia. Additionally, a bidirectional causal relationship between the two variables was found.

The natural resources and CO2 emissions in the 16 European economies were shown to have a long-term, substantial positive relationship, according to Bekun et al. (2019). It implies that if abatement and rehabilitation actions are neglected, excessive dependence on natural resources will affect their environmental lifetime. Ulucak and Ozcan (2020) assessed the impact of natural resources on the environmental performance of OECD economies. They concluded that increasing emission levels were caused by natural resource extraction. Nevertheless, Balsalobre-Lorente et al. (2018) found a link between natural resources and CO2 emissions in France, Italy, Germany, Spain, and the UK from 1985 to 2016. The empirical findings showed that natural resources improve the environment's performance. The effect of natural resource exploitation on environmental quality in South Asian nations was examined by Xue et al. (2021) during the years 1991–2018. They addressed cross-sectional dependency by using dynamic common correlated effects. The findings of the long-run estimation show that natural resources have a negative correlation with the ecological footprint and a positive and substantial correlation with all greenhouse gas emissions.

The above literature provides the order of the influences of natural resources and human capital that did not hold through fully updated data availability. To realize this gap, current research will be apprehended to stop this flaw and check how human capital and natural resources can produce a green atmosphere in five South Asian regions (Pakistan, India, Sri Lanka, Nepal, and Bangladesh). In this way, the study extracts the data from 1975 to 2021 to check the effects of human and natural capital, which are taking place for ecological excellence. Moreover, previous research regarding the region of South Asia didn't examine the combined consequences of human capital and natural resources on EF using their interaction term, so this study is casting off human capital and natural resource rents by taking their interaction term.

3 Data, model, and methodology

3.1 Data description

This research study is conducted with a panel of five South Asian states (Pakistan, India, Sri Lanka, Nepal, and Bangladesh), excluding Afghanistan, Bhutan, and Maldives due to the unavailability of the data. The current study will involve ecological footprint, human capital, natural resource rents, GDP, the square of GDP, urban population, and industrial value-added. By casting off forty-seven years' time span 1975–2021. Table 1 contains the data description with their respective sources, and the estimate flowchart used in this work is shown in Fig. 2. Using these parameters mentioned above as our attention, we created the following ecological footprint model:

Table 1 Data Description
Full size table
Fig. 2
figure 2

Flow Chart of the Analysis

Full size image
$${\text{EF}}=f\left(\mathrm{ HC},\mathrm{ GDP}, {{\text{GDP}}}^{2},\mathrm{ NRR},\mathrm{ URP},\mathrm{ INV}\right)$$
(1)

The above equation displays that EF denotes the ecological footprint, NRR is natural resource rents, HC shows human capital, GDP is a gross domestic product, GDP2 is growth square, URP is urban population growth, and INV is the industrial value added.

$${{\text{EF}}}_{{\text{it}}}={\upbeta }_{0}+{\upbeta }_{1}{\text{lnHC}}+{\upbeta }_{2}{{\text{GDP}}}_{{\text{it}}}+{\upbeta }_{3}{{GDP}^{2}}_{{\text{it}}}+ {{\text{B}}}_{4}{{\text{lnNRR}}}_{{\text{it}}}+ {{\text{B}}}_{5}{{\text{lnURP}}}_{{\text{it}}}+ {\upbeta }_{6}{{\text{lnINV}}}_{{\text{it}}}++{{\text{u}}}_{{\text{it}}}$$
(2)

where \(u\) is the error term, and ln is the natural log. Countries and times are denoted by the subscripts i and t, accordingly.

The novelty of the study also rests in the fact that it evaluates the moderating influence that HC has on the link between NRR and EF in studied countries. We investigate the moderating influence of HC, which is connected to NRR, to determine the effect of interaction that is exerted on EF. The NRR is known to be improved by HC (Jahanger et al., 2023). Consequently, it is reasonable to assume that it will raise the level of environmental sustainability across all of the countries that were investigated. In the extended Eq. (2), the interaction term between HC and NRR (HC*NRR) is added. The empirical model that is stated in Eq. (3) is utilized to investigate the moderating influence that HC*NRR has on academic performance.

$${{\text{EF}}}_{{\text{it}}}={\upbeta }_{0}+{\upbeta }_{1}{\text{lnHC}}+{\upbeta }_{2}{{\text{y}}}_{{\text{it}}}+{\upbeta }_{3}{{y}^{2}}_{{\text{it}}}+ {{\text{B}}}_{4}{{\text{lnNRR}}}_{{\text{it}}}+ {{\text{B}}}_{5}{{\text{lnURP}}}_{{\text{it}}}+ {\upbeta }_{6}{{\text{lnINV}}}_{{\text{it}}}+ {\upbeta }_{7}{\text{ln}}{{\text{HC}}*{\text{NRR}}}_{{\text{it}}}+{{\text{u}}}_{{\text{it}}}$$
(3)

4 Methodology

4.1 Cross-sectional dependency

South Asian countries are economically and geographically linked with each other. Therefore, if a country experiences a shock, it can upshot the economy of other republics. Thus, it is imperative to check the cross-sectional dependence (CSD) among variables before the main estimation of the analysis. So, this study uses the CSD test for models. To quantify CSD in the panel data, we employed the CSD test created by Pesaran (2021). In panel data analysis, inspecting CSD is essential to avoid biased or deceptive conclusions.

4.1.1 Panel unit root test

Next, the variables' integration order is investigated. First-generation unit root techniques such as Levin-Lin and Chu and Im, Pesaran, and Shin (IPS) cannot alleviate the issue with CSD (Ali et al., 2023; Pesaran, 2007). According to Pesaran (2007), cross-sectional augmented Dickey-Fuller (CADF) and cross-sectional augmented IPS (CIPS) unit root tests and their projections are preferable to the first generation when applied to a CSD situation. The CADF and CIPS tests are evaluated based on the null hypothesis, which states that the elements are not stationary despite the alternative hypothesis used to evaluate the test. As a result, we used the CADF and CIPS unit root tests while considering CSD (Pesaran, 2007).

Here is the formula for the test statistic:

$$\Delta CA_{i,t} = \phi_i + \phi_i Z_{i,t - 1} + \phi_i {\mathop {CA}\limits^{\_\_\_\_\_}}_{t - 1} + \sum_{l = 0}^p {\phi_{il} } \Delta {\mathop {CA}\limits^{\_\_\_\_\_}}_{t - 1} + \sum_{l = 0}^p {\phi_{il} } \Delta CA_{i,t - 1}$$
(4)

where \({\mathop {CA}\limits^{\_\_\_\_\_}}_{t - 1}\) and \(\Delta {\mathop {CA}\limits^{\_\_\_\_\_}}_{t - 1}\) are the cross-section averages. The statistics of the CIPS test are:

$${\mathop {CIPS}\limits^\wedge } = \frac{1}{N}\sum_{i = 1}^n {CADF_i }$$
(5)

4.1.2 Panel cointegration

To increase power and frequently include a two-stage process. The first phase is intensely checking the panel unit roots test and panel cointegration assessment, which are cast off. Another phase is the cointegration test on board. The study outfits the Pedroni Kao and Westerlund cointegration test for the presence of cross-sectional dependency if variables are combined into order one. It permits the short-run and long-run dynamics to fluctuate, so examination is considered superior to other cointegration tests. This cointegration test practices info more efficiently than residual founded investigations and doesn't execute any mutual dynamic limitation on the data. Suppose a long-term co-integrating relationship happens among human resources, natural resources, and ecological deprivation.

The panel cointegration is an excellent method for cointegration where the countries N ≥ 2. The cointegration test is used when the variables are stationary. The panel unit root test and panel cointegration test are intensification examinations of increments in power and often contain a two-step technique. The first step is to check the panel unit's wiring. The other step is the cointegration test in the panel. The panel cointegration is an excellent technique for cointegration where the countries N ≥ 2. The cointegration test is used when the variables are stationary. Pedroni (1997, 1999, 2004) introduced the cointegration method, which is a panel based on residuals that consider the heterogeneity of a particular between effects, slope coefficients, and individual linear trends countries, Pedroni (2004) uses the following regression:

$${{\text{y}}}_{{\text{it}}}={{\text{a}}}_{{\text{i}}}+{\updelta }_{{\text{i}}}{\text{t}}+{\upbeta }_{{\text{i}}}{{\text{X}}}_{{\text{it}}}+{\upmu }_{{\text{it}}}$$
(6)

The second technique for the cointegration test is the Kao test, which was developed by Kao (1999) and performs homogeneity in a panel. This study reduces rationality between natural resources, human capital, and environmental degradation in South Asian countries. The study tested the invalid number assumption, and cointegration against configuration was considered a variable. This study reduces rationality among natural capital, human capital, and ecological poverty in South Asian states. The research tested the invalid number assumption, and cointegration against configuration was considered a variable.

Therefore, for the panel data test introduced by Westerlund (2007), research uses the cointegration test by considering operational dynamics as relative to residual dynamics. Therefore, the research does not impose any type of limitations on any common factor. Additionally, the error correction model (ECM) is assumed by Westerlund (2007). If altogether variables are linked to mandate one or I(1) and so on, it can be engraved as per:

$$\Delta {{\text{x}}}_{{\text{it}}}={\uptheta }_{{\text{i}}}{{\text{d}}}_{{\text{i}}}+{\uppi }_{{\text{i}}}\left({{\text{x}}}_{{\text{it}}-1}-{\upbeta }_{{\text{i}}}{{\text{Y}}}_{{\text{it}}-1}\right)+\sum_{{\text{j}}=1}^{{\text{m}}}\mathrm{\pi ij}{\Delta {\text{x}}}_{{\text{it}}-{\text{j}}}+\sum_{{\text{j}}=1}^{{\text{m}}}\mathrm{\pi ij}{\mathrm{\varphi }}_{{\text{ij}}}{\Delta {\text{x}}}_{{\text{it}}-{\text{j}}}+{{\text{u}}}_{{\text{it}}}$$
(7)

The above equation can be made using the ordinary least square method written as:

$$\Delta {{\text{x}}}_{{\text{it}}}={\uptheta }_{{\text{i}}}{{\text{d}}}_{{\text{i}}}+{\uppi }_{{\text{i}}}\left({{\text{x}}}_{{\text{it}}-1}-\uptau {{\text{Y}}}_{{\text{it}}-1}\right)+\sum_{{\text{j}}=1}^{{\text{m}}}\mathrm{\pi ij}{\Delta {\text{x}}}_{{\text{it}}-{\text{j}}}+\sum_{{\text{j}}=1}^{{\text{m}}}\mathrm{\pi ij}{\mathrm{\varphi }}_{{\text{ij}}}{\Delta {\text{x}}}_{{\text{it}}-{\text{j}}}+{{\text{u}}}_{{\text{it}}}$$
(8)

whereas πi designates the speediness of adjustment to control. The overhead equation authorizes this back in balance by the imposition of arbitrary \(\tau\) that the coefficient πi remains genuine.

4.1.3 Panel ARDL

The research study used panel autoregressive distributed lags (ARDL) or pooled mean group (PMG) technique for the long-run and short-run analysis, which was presented by Pesaran and Smith (1995) and Pesaran et al. (1999). In addition to an approximation of long-term and short-term connotations amongst selected variables, the panel is ARDL, and the PMG technique is used. This technique also tests the vigorous heterogeneous problem in cross-sections, despite guessing the relationship in variables, which can be transcribed as a long-term connotation amongst variables. Panel ARDL or PMG model is written as follows:

$${{\text{lny}}}_{{\text{it}}}=\sum_{{\text{j}}={\text{i}}}^{{\text{p}}}{\mathrm{\alpha }}_{{\text{ij}}}{{\text{lnY}}}_{{\text{it}}-1}+\sum_{{\text{j}}=1}^{{\text{q}}}\mathrm{\beta ij}{{\text{lnX}}}_{{\text{it}}-{\text{j}}}+{\mathrm{\varphi }}_{{\text{i}}}+{{\text{u}}}_{{\text{it}}}$$
(9)

whereas, lnYit is a dependent variable and lnXit shows independent parameters. Here, \({\alpha }_{ij}\) is the intercept term, and \(\sigma\) is the coefficient of explained variable lagged values. Moreover, \({\varepsilon }_{it}\) is an error term, and \(\varphi\) is group effect. One may determine the short-run dynamics of the panel ARDL specification by creating an error correction model in the following form:

$$\Delta {{\text{lny}}}_{{\text{it}}}=\mathrm{\varphi }{{\text{ECT}}}_{{\text{it}}}+\sum_{{\text{j}}={\text{i}}}^{{\text{p}}}{\mathrm{\alpha }}_{{\text{ij}}}^{*}{{\text{lny}}}_{{\text{it}}-{\text{j}}}+\sum_{{\text{j}}=0}^{{\text{q}}}{\upbeta }_{{\text{ij}}}^{*}\Delta {{\text{lnx}}}_{{\text{it}}-{\text{j}}}+{{\text{u}}}_{{\text{i}}}+{{\text{u}}}_{{\text{it}}}$$
(10)

Here, \(\varphi\) represents the speed of adjustment toward the long-run equilibrium. Besides, \({ECT}_{it}\) is the error correction term.

5 Results and discussion

Before going to the panel data econometric investigation, a detailed descriptive analysis is brought out of 235 observations. In Table 2, the descriptive statistics of regressors are stated. For instance, the trends are represented by the mean values, which show whether the variables are rising or declining. Except for the NRR value, all other variables' values are positive. The mean values of EF, GDP, GDP2, HC, INV, NRR, and URP are 1.05, 4.59, 28.24, 1.50, 28.0, -0.003, and 0.83, correspondingly. Furthermore, the standard deviation shows the difference between the observed value and the mean at a specific point in time. The standard deviation of EF, GDP, GDP2, HC, INV, NRR, and URP are 1.08, 2.67, 20.40, 0.21, 1.70, 0.95, and 0.87. Additionally, the maximum and minimum values show the variable's range, where the earlier denotes the variable's highest value within the chosen time, and the latter reflects its lowest value. The maximum values of EF, GDP, GDP2, HC, INV, NRR, and URP are 9.35, 9.68, 101.5, 2.11, 31.36, 2.21, and 2.38. The minimum values of EF, GDP, GDP2, HC, INV, NRR, and URP are 0.45, −10.07, 0.01, 1.16, 23.7, −2.66, and −3.04. Table 3 also shows the correlation matrix of the studied variables. Similarly, Fig. 3 depicts the trends of the variables used in this study throughout 1975–2021.

Table 2 Descriptive Statistics
Full size table
Table 3 Correlation Matrix
Full size table
Fig. 3
figure 3

Trends of the variables incorporated in the analysis

Full size image

Where all units in the same cross-section are correlated, panel data may be the theme of broad cross-sectional dependence (CSD). To influence approximately unnoticed general aspects, though possibly in dissimilar ways, which are corporate to all units and affect each other, this is usually credited. Thus, the research checks the cross-sectional dependency expending to observe whether residuals are connected in crossway states or not, employing the CSD test. The outcomes are stated in Table 4 below.

Table 4 Cross-sectional dependency test results for each series
Full size table

Likewise, after confirming the CSD in the panel, we employed the second-generation unit root tests, which are shown in Table 5. Consequently, this study used the second-generation panel unit root test, the cross-sectional ADF, and the CIPS test of Pesaran (2007). These tests are more reliable in the presence of CSD. EF, HC, NRR, and INV are not stationary at level, but these are stationary at the first difference. However, GDP, GDP2, and URP are stationary at the level.

Table 5 Panel unit root test results
Full size table

Further, the study examines the presence of a cointegration relationship using Westerlund's (2007) cointegration test. Because the model suffers from cross-sectional dependency, the research employed a second-generation Westerlund (2007) cointegration test. The cointegration links between the dependent variable and the independent variables have been checked before running the panel ARDL approach. However, we employed three kinds of panel cointegration tests in this research. That is, the Pedroni (1999, 2004), Kao (1999), and Westerlund (2007) tests. Table 6 displays the outcomes of the Pedroni test. The outcomes show that all values are significant at the 1% and 5% levels, respectively. Table 6 indicates that within the dimension, the statistics with accurate probability values show the cointegration relationship among variables of both models. Tables 7 and 8 summarize the effects of Kao's integration test and the Westerlund test, which rejects the null hypothesis of no cointegration, representing a long-run cointegration affiliation amongst variables. The Kao cointegration analysis elaborates on the existence of cointegration. Besides, Table 8 Westerlund test rejects the null hypothesis about no cointegration among statistics, indicating that a long-run cointegration association exists among variables.

Table 6 Pedroni cointegration test results
Full size table
Table 7 Kao's cointegration test results
Full size table
Table 8 Westerlund cointegration test results
Full size table

To investigate the long-run association of regressors, we used the panel ARDL test. The findings of panel ARDL are presented in Table 9. It is a decision rule that when the long-run affiliation exists, the model is significant and indicates the presence of a long-run relationship. The results indicate that HC is positively concerned with environmental pollution and is significant. The results based on human capital consequences explain that human capital growth (short and long run) produces an important increase in the EF. A 1% increase in HC will increase EF by 0.13% in the long run and 0.086% in the short run. On the other hand, this should not come as a surprise because HC plays a significant role in the expansion of the economy. Higher levels of HC are associated with increased economic activity, which, according to the EKC hypothesis, tends to add to the pollution levels. It has been argued in previous research that environmental pollution is directly tied to human economic activities as a result of rising energy demand and usage (Haini, 2021). These findings are consistent with the conclusions that have been drawn from previous research.

Table 9 Panel ARDL Results
Full size table

As seen in Table 9, NRR has a negative and significant effect on EF. When taking into account the statistically significant finding, it can be concluded that increasing the utilization of NRR leads to a 0.10 and 0.09% improvement in the performance of the environment over the long and short run, respectively. The outcomes of this study indicate that the countries that were researched employ environmentally friendly and risk-free methods to acquire NRR, which in turn helps to reduce the quantity of pollution that is released into the atmosphere. Similarly, this evidence suggests that using strategies for the efficient use of NRR lowers EF, improving environmental sustainability. The inverse connection suggests that economies in South Asia are using strategies to efficiently manage the use of NRR to achieve sustainability. This suggests that the consequences of natural resources' rent-seeking seem to be completely safe for the climate, leading to the pollution-prevent consequences that the results show. This makes sense since the profits from NRR may also be utilized as incentives to keep improving the environment's efficiency over time in the economies under study without degrading the quality of the ecosystem. The outcomes correlate with Balsalobre-Lorente et al.'s (2018) findings.

Table 9's empirical results show that GDP has a significant and favourable impact on EF. More precisely, in the long and short run, ecological sustainability is decreased by 0.10 and 0.008%, respectively, for every 1% increase in GDP. The integrity of the ecosystem tends to deteriorate until it crosses the threshold. Ecological issues and sustainability concerns are being raised by the middle-income stage of development that the South Asian countries are going through. These findings are consistent with earlier research by Dogan and Shah (2021) and Qayyum et al., (2023). Furthermore, the validity of EKC in South Asian countries across the research period is confirmed by the GDP2's negative value. A 1% increase in GDP2 reduces the EF in the long run by −0.001% and in the short run by −0.0004% over time, based on the negative coefficient of GDP2. It demonstrates that once the South Asian economies have reached a certain level of economic growth, they may improve ecological sustainability. Adopting green technologies, embracing ecological manufacturing, and replacing high-polluting industries with low-carbon ones all contribute to an optimal economic environment. Previous studies (Ali et al., 2022a, 2022b; Gyamfi et al., 2022; Rout et al., 2022) support the application of the EKC theory to South Asian countries.

The coefficient of URP, which is notably negatively significant, implies that South Asia's EF is lessened by URP. Researchers who claim that URP benefits ecological deterioration conflict with this finding (Al-Mulali & Ozturk, 2015; Charfeddine, 2017; Rashid et al., 2018). Our results, however, are consistent with those of Sharma (2011) and Charfeddine and Ben Khediri (2016). A 1% rise in URP reduces the EF in the long run by −0.122% and in the short run by −0.013% over time. URP has the potential to decrease EF in several ways. Through the realization of economies of scale, URP may raise productivity and resource conservation. The service industry, which has less of an impact on the atmosphere, may benefit from it. Public services like waste management, water supply, and sanitation are more affordable to build and run in metropolitan areas. Moreover, URP promotes energy effectiveness, green technology, and inventiveness (Charfeddine & Mrabet, 2017).

According to INV's findings, a 1% boost in INV raises EF in the long run by 0.083% however, it decreases EF in the short run by −1.83%, respectively. The long-run finding is not unexpected at all. Industrial activities have been on an upward trajectory in recent years and the majority of production operations involve the use of fuel for everyday tasks, which results in increasing emissions. Furthermore, places where mining operations make up a significant share of GDP are big contributors to the EF. As a result of the transformation that industrialization brings about in mining and manufacturing activities, the conclusion is even more convincing. These results are in line with the findings of Raheem and Ogebe (2017) for 20 African nations and Liu and Bae (2018) in China.

Toward examining the combined effect of HC*NRR on environmental quality in South Asia, the research explores that HC*NRR have a confident association with environmental quality. The elasticity of HC*NRR is 1.75 in the long run, while in the short run 1.50, which indicates that if there is a one per cent increase in the HC*NRR, then there would be a 1.75% increase in EF in South Asia. It means NRR and the implication of HC may lead the environmental pollution. They both have their own influence on the environment, but the combined effect of HC and NRR has a bigger impact on the ecosystem. By doing more than simply globalizing the economy, it would be important to advocate an agenda that is favourable to the environment and boosts sustainable development, both of which may stimulate the use of NRR and HC throughout studied nations. HC can reduce the destructive influences of NRR on the environmental footprint. Hence, it demands human capital expansion (Nathenial, 2021).

6 Conclusion

The present study examines the measurable factors in five South Asian countries (Pakistan, India, Sri Lanka, Nepal, and Bangladesh) and uses panel data to explore further possibilities by examining the impact of natural resource rent and human capital on ecological degradation. For empirical analysis, cross-sectional dependency, second generation unit root test CIPS and CADF test, Pedroni, Kao and Westerlund test for cointegration analysis, and panel ARDL test for the long and short run results are employed. The outcome defines that human capital has a negative and significant alliance with environmental quality, whereas economic growth and economic growth square affect environmental quality positively and negatively confirming the existence of the EKC hypothesis. Urban population growth and natural resources have significant negative associations with environmental pollution, whereas industrial value added increases the pollution levels in South Asia.

6.1 Policy implications

In practice, this study suggests policies that might assist South Asian economies in accomplishing sustainable development goals and improving environmental quality.

  • As the study findings show a negative effect of human capital on ecological footprint, the government needs to put more budget into initiatives aimed at developing human capital. For instance, South Asian nations should start implementing ecological awareness campaigns to educate the population about climate change and the value of pro-environmental behaviours like recycling, using renewable energy sources, conserving energy and water, and preserving energy.

  • As the use of natural resources is beneficial for the environment, policymakers must focus on increasing the stock of natural resources, monitoring their decline, and other concerns such as forest fires and exploitation of natural resources. Expanding greenery, regulating contamination and ecological decline, and promoting reduce-reuse-recycle campaigns can assist in slowing the decline of natural resources.

  • South Asian countries are hastening the approach of the EKC tipping point. To achieve its yearly growth target, South Asian economic progress should collaborate with the green economy or green growth program.

  • Our findings suggest that urbanization has an unfavourable impact on pollution; however, the present urbanization rate is gradually increasing in South Asian countries. Urban development must prioritize rising urban density and infrastructural expansion, and compact cities should be constructed to maximize resource efficiency.

  • Lastly, the governments should also impose ecological taxes to discourage activities contributing to ecological humiliation by the industrial sector, as we found a positive impact of industrial value added on ecological footprint. Governments focus on green technology advancement to move from non-renewable to renewable energy sources and accelerate economic growth.

6.2 Limitations and future research suggestions

Nonetheless, this research offers important empirical information about the impact of natural resources and human capital on the ecological footprint of the South Asian economies. However, this study is constrained in a few aspects. This research is particularly restricted in terms of determining total natural resources and their influence on sustainable development. Future studies should incorporate other natural resources such as stone, metals, sand, air, water, etc. Future studies might expand on the current findings by looking at the impact of fossil fuel and renewable energy usage on the prospective progress of these countries. Furthermore, future studies may increase the time span and cross-sections.