This paper examines whether India’s growth–environment trade-off is sustainable. It has two specific objectives. The first is to investigate the long-run co-integrating relationship between the ecological footprint (a proxy variable for the environment) and GDP per capita (a proxy variable for economic growth), with urbanisation, trade and biocapacity as additional variables. The second objective is to examine the applicability of the ‘N’-shaped Environmental Kuznets Curve (EKC) hypothesis in the Indian context. To accomplish these objectives, the study considers the period from 1971 to 2017 (46 years) and sources its data from the World Development Indicators and the Global Footprint Network. The methodology employed to analyse the relationship between the variables is ARDL cointegration and Toda-Yamamoto causality, while two variants of GDP (square and cubic) are used to validate the ‘N’-shaped EKC in the Indian context. The results revealed a strong long-term relationship (cointegration) between the variables. It was observed that all the variables except trade impacted the ecological footprint, with causality flowing from all the variables towards the footprint. However, an ‘N’-shaped EKC was not proven for India. These results have important implications. The long-run lagged error adjustment coefficient (ECM(-1)) revealed a slow adjustment process towards achieving sustainable development targets, which indicates that India’s environmental compromise in pursuit of economic growth still persists. While the nationwide initiatives, such as the National Solar Mission and the FAME schemes, are praiseworthy, the study recommends much more stringent policy measures for India. Stricter policies may require moderating some of the country’s growth targets in order to achieve transition towards an environmentally sustainable development path well aligned with the Sustainable Development Goals (SDGs) and COP26 commitments.

EKC, ARDL, Causality, ‘Ecological Footprint’, Sustainable Development

The contribution of the ecological system to the existence and nurturing of humanity is unprecedented, well documented and beyond debate. For centuries, the human race has existed in harmony with the ecological system. However, the pursuit of a better quality of life and improved living conditions over the last few decades has disturbed the ecological system’s delicate balance, resulting in rising global temperatures, melting glaciers, rapid extinction of wildlife, unpredictable rainfall and other climate-related changes (Hassan et al., 2019; Shahbaz et al., 2018).

The issue of global ecological imbalance has been widely discussed at many national and international forums and summits. At these events, developing countries have often highlighted the role played by advanced economies in contributing to environmental degradation. These discussions tend to emphasise that the development patterns and policies adopted by advanced economies in the past were primarily aimed at achieving higher growth and income levels. However, these policies have also contributed to ecological degradation and imbalances within fragile ecosystems.

In debates about environmental degradation at various forums, representatives of developing economies often question the rationale behind imposing stricter sustainability requirements on them when such requirements were not imposed on developed economies during their own growth cycles within the broader context of sustainable development. Developed countries generally respond by pointing out that the economic and environmental conditions during their own growth phases differed considerably from the present day. At the same time, however, these nations are willing to support developing countries in global efforts to reduce emissions and promote environmental sustainability (Allan, et al., 2023; Agarwal & Narain, 2012; Stern, 2008).

According to a study conducted by the Centre for Global Development, which analysed historical carbon emissions data over 160 years from 1850 to 2011, the contribution of developed economies to carbon emissions was found to be approximately 79% (www.cgdev.org).

With regard to the current emissions scenario, the worrying aspect is that the share of emissions from developing countries has already surpassed that from developed countries (Sharma et al., 2021). In their desire to achieve higher growth rates, developing economies are often found to be compromising environmental concerns, thereby raising environmental degradation to unsustainable levels. The main issue is the extremely high consumption of non-renewable fuels in developing economies, whereas developed countries have substantially replaced this with renewables. According to estimates, traditional fossil fuels contribute 68% of GHG emissions in developing economies (Hassan et al., 2019). Furthermore, rapid urbanisation in these economies has led to a disproportionate distribution of emissions, with urban areas being the main contributors.

Some studies have been conducted on the potential future impact of environmental degradation, and the results reveal an inverse relationship between the level of economic development and climate change impact. This implies that developing economies, due to their lower level of development, are likely to experience a greater impact of climate change, primarily due to their limited capacity for environmental adaptation (Nkengfack & Fotio, 2019). Nevertheless, many developing countries have initiated energy-efficient strategies or fine-tuned existing policies to target greenhouse gas (GHG) emissions.

While appreciating such steps, environmentalists add a note of caution, stating that such a strategy can only serve as a temporary solution and that countries ultimately have little choice but to switch to renewable energy sources. According to them, the goal of any economy must be to transition from non-renewable to renewable energy, coupled with strict environmental regulation enforcement. Environmentalists also caution that any leniency in enforcement would render the transition meaningless. Research has revealed that countries where environmental regulations are compromised over time tend to become pollution havens, attracting dirty industries and thereby importing more carbon. It is worth noting that cross-border energy transactions have been one of the primary reasons for exceptional growth in some developing economies. Therefore, any reduction in this area may be difficult to implement, as it would mean cutting back on growth targets, which most developing countries are unwilling to do (Shahani et al., 2021; Sharvini et al., 2018; Ahmed et al., 2022; Murshed et al., 2021).

Moving on to empirical research in the field of growth and environment, this area of study gained momentum following the introduction of the renowned Environmental Kuznets Curve (EKC) by Grossman and Krueger (1995). This theory suggests that developed countries are less polluting than developing and underdeveloped countries, and that a reduction in environmental degradation is linked to income levels. According to EKC, once a country enters the ’threshold’ income zone, environmental degradation tends to reverse, i.e. the upward rise in the EKC curve is halted and the graph reaches a turning point (Turning Point 1, Fig. 1). Beyond this turning point, the curve starts to fall, creating an inverted U-shape.

Fig. 1. 

‘N’ shaped EKC

On the other hand, as incomes continue to increase, a stage is reached where environmental degradation is at its maximum (i.e. the EKC curve reaches its minimum), and the curve begins to rise again as incomes increase further. Researchers have termed this an ‘N’-shaped EKC curve. Thus, a typical ‘N’-shaped curve (Fig. 1) has two turning points. The first turning point creates an ‘inverse U’-shaped curve, while the second turning point extends this to an ‘N’-shaped curve. It is this ‘N’-shaped curve that is more relevant today for most economies.

In order to validate EKC, researchers employed many proxy variables for ‘environmental degradation’ in their studies, including CO2 emissions, SO2 emissions or PM10 emissions. Income or GDP is usually used as a standard proxy for ‘growth’. On the other hand, some economists challenged emission indicators because they are unable to comprehensively capture environmental degradation and recommended new variables such as ‘ecological footprint’, which can measure degradation quantitatively.

The new indicator, the ‘ecological footprint’, includes the effects of all the goods and services produced, as well as the waste generated, and is considered to be a superior representation of environmental degradation. In simple terms, the variable incorporates all the activities required to support a given lifestyle (Rashid , et al., 2018; Global Footprint Network, 2014). Due to its direct and indirect focus on environmental impact, the term has gained importance, replacing most other proxies for environmental degradation. Recent studies, such as those by Mrabet and Alsamara (2016), Mrabet et al. (2017), Öztürk et al. (2016) and Wang et al. (2013), have considered the variable ‘ecological footprint’ to be their proxy of choice for environmental degradation.

Some studies that considered the ecological footprint as a proxy variable found mixed results regarding its relationship with economic growth and the validity of the EKC. While studies by Ulucak and Bilgili (2018), Öztürk et al. (2016) and some others found the EKC to be valid, Destek et al. (2018) and Bello et al. (2018) were unable to establish the hypothesis. Then, in addition to growth, the ‘ecological footprint’ indicator was tested against other variables such as urbanisation, foreign direct investment (FDI), financial development and environmental technology. Some of these variables have also been included in the present study, with a detailed discussion of their relationship to the ecological footprint covered in the literature review section.

Another variable that is becoming increasingly popular in empirical environmental studies, often alongside the ecological footprint, is ‘bio-capacity’. According to the Global Footprint Network (GFN), all commodities and services carry an embedded amount of bioproduction area (land and sea, including associated waste), which is termed ‘bio-capacity’ and expressed in global hectares. If a country’s bio-capacity exceeds its ecological footprint, it is said to have an ecological reserve. Conversely, if the ecological footprint exceeds the bio-capacity, the country faces an ecological deficit. According to data provided by GFN, more than 80% of the world’s population resides in countries with an ecological deficit (www.footprintnetwork.org ). Furthermore, it is estimated that the world today requires the equivalent of 1.5 times the total surface area of the Earth to sustain the current ecological footprint (Hassan et al., 2019).

In light of the above discussion and bearing in mind the objectives of the study, it was decided to explore whether the path envisaged for India’s growth story is sustainable. To answer this question, the study aims to establish a relationship between ‘ecological footprint’ and a group of growth- and development-related variables, which have been conveniently classified as ‘traditional’ and ‘non-traditional’. The former includes commonly used single-data-type variables, while the latter considers two complex variables: ‘ecological footprint’ and bio-capacity. The study would also test the validity of the ‘N’-shaped EKC model in the Indian context, and attempt to determine whether a long-term relationship exists between the ecological footprint and these variables. The behaviour of each independent variable towards ‘ecological footprint’ is further investigated to determine whether it is consistent with expectations. Furthermore, as India aims to become a developed economy by 2047, the study’s unique contribution will be to determine if this can be achieved in a sustainable manner. With these considerations in mind, we move on to our next section, which is the literature review.

In the literature review, we examine some studies that have considered the ‘ecological footprint’, which is the environmental proxy used in the present research. First, we examine recent papers that have tested for the EKC relationship, followed by empirical studies on co-integration and other methodologies involving the variable ‘ecological footprint’.

Nathaniel et al. (2021) tested for EKC amongst MENA countries by taking ‘ecological footprint’ and GDP. The results revealed that the EKC hypothesis was applicable during the early phase of economic development. Furthermore, environmental degradation was clearly visible during this early period; however, with more economic development, this degradation was seen to reverse, thereby confirming the validity of the EKC model. Subsequently, Sharma et al. (2021) conducted a similar investigation in eight South Asian countries. The results revealed two turnarounds in the EKC curve: the first was in line with the EKC model, while the second was above the expected range, indicating that future economic growth in these countries may be unsustainable. However, the results did prove the shape to be in line with the ‘N’-shaped EKC model.

In a large sample of 85 countries divided into three categories (high, medium and low income), Numan et al. (2022) tested the ‘N’-shaped EKC model. The results were mixed, with only a few countries supporting the hypothesis. The researchers therefore ruled out generalising the ‘N’-shaped EKC and recommended customised environmental policies for each country. On the other hand, Farooq et al. (2023) tested the validity of the EKC curve in India to determine the country’s stage of development by considering six disaggregated footprints, as defined by the term ‘ecological footprint’. The results revealed that five of the six footprints validated the EKC. The study projected that India’s environmental degradation would become sustainable by 2034; however, it also warned of considerable depletion of grazing and forest land before this could be achieved.

Thus, the viewpoint regarding the ‘ecological footprint’ as an environmental proxy appears to be no different to that of other environmental proxies, such as CO₂ and SO₂, which are also considered by researchers whose studies produce mixed results regarding the validity of the EKC. In other words, some studies report its validity, while others do not.

Apart from validating the EKC, the ‘ecological footprint’ variable has been tested for co-integration, causality and other relationships against many variables, including environmental regulations, renewable and non-renewable energy variables, financial development, foreign direct investment (FDI), urbanisation, democracy and life expectancy, among others. Studies that have considered ‘environmental regulations’ (using environmental patents as a proxy) include those by Danish et al. (2020), Ahmed et al. (2022) and Fotis and Polemis (2018). All of these studies report that ‘environmental regulations’ not only decrease the ecological footprint directly, but also stimulate ‘renewables’, which are known to reduce the ecological footprint. The biggest hurdle identified with respect to ‘environmental regulations’ was that they were not decided by market forces, but by political institutions that manage state affairs, thus paving the way for tight environmental regulations in a few countries and lax regulations in others.

Then, the variable ‘growth’, proxied by GDP, was found to contribute to ‘environmental degradation’ and enjoy a positive relationship with ‘ecological footprint’ due to its connection with the mass consumption of resources. Similarly, life expectancy was also found to be positively related to ecological footprint, with urban life expectancy contributing more detrimentally than rural life expectancy. Age was studied separately, despite being closely linked to life expectancy. Studies on ‘age’ have revealed that environmental awareness increases with age and is expected to result in lower emissions. However, this has not been observed in empirical studies, and it has been found that there is a positive relationship between ‘age’ and ‘ecological footprint’, rather than a zero or negative relationship (Zagheni, 2011). Among the other variables, global trade was found to impact the ecological footprint both positively and negatively. If trade induces the flow of energy-intensive technologies into a country, this is expected to increase its ecological footprint. However, trade may also bring green technologies through technological substitution, which would have a negative impact on the ecological footprint (Murshed et al., 2021; Ahmed et al., 2020).

One variable that has been comprehensively studied in relation to the ‘ecological footprint’ is renewable energy. Most studies, including those by Sharif et al. (2020), Nathaniel et al. (2020a, 2020b) and Radmehr et al. (2022), have found a negative correlation. Nathaniel et al. (2020a) revealed a negative relationship for MENA countries, whereas Sharif et al. (2020) obtained similar results for Turkey.

Some studies have attempted to establish a link between ‘ecological footprint’ and variables such as ‘FDI’ and ‘financial development’. With respect to foreign direct investment (FDI), Solarin and Al-Mulali (2018) found that this variable had no impact on the ecological footprint. However, Baloch et al. (2019) and Mrabet et al. (2017) concluded that financial development increased the ecological footprint. Regarding the complex variable ‘bio-capacity’, Hassan et al. (2019) demonstrated that ‘bio-capacity’ was statistically significant and negative, indicating that the ‘bio-capacity’ accumulated was detrimental to the ecological footprint.

Based on the above review of the impact of different variables on the ‘ecological footprint’, we can broadly conclude that these results are in line with expectations and are no different to those of other environmental proxy variables such as CO₂, SO₂, PM₁₀ and other emissions considered by researchers in the past. Furthermore, considering the nature of the variables that researchers have considered in different studies, we conclude that these can conveniently be divided into traditional variables (e.g. growth, trade, urbanisation, renewable and non-renewable energy) and non-traditional variables (e.g. democracy, environmental regulations, bio-capacity). Our literature review suggests that the relationship between non-traditional variables and ‘ecological footprint’ is stronger than that between traditional variables and ‘ecological footprint’.

In light of the above, our study has two objectives: first, to test the ‘N’-shaped EKC model for India; and second, to test the long-run co-integrating relationship between the ecological footprint and a mix of traditional and non-traditional variables. The co-integration model applied is the ARDL model, chosen because of the difference in the level of integration among the included variables (see Section 4: Research Methodology for further details).

Table I below summarizes information about the five variables considered in the study. All variables are log-transformed; the sample period covers 46 years of annual data from 1971-2017.

Apart from these five variables, we also included two transformations of the variable GDP, namely the square and the cube of the GDP variable, which facilitate the testing of the N-shaped EKC hypothesis. According to Baloch et al. (2022), the three turns as seen in N shaped EKC model represent scale, composition and technique obsolescence effects. If we consider the relationship in the expected way, then income, income squared and cubic income (i.e. GDP and its two transformations) should have a direct, then indirect and then direct relationship with the ecological footprint in order to validate an N-shaped relationship. According to proponents of the ‘N’-shaped curve, the composition effect, which is directly a function of income, sets in during the initial stages of economic development. This may change or reverse later for some economies that use excessive energy resources or fail to replace old, energy-intensive equipment with newer, cleaner alternatives. The significant positive income coefficient indicates that the country may still be using obsolete, energy-intensive technology.

Concerning the variable ‘bio-capacity’, it is expected to be negatively related to the ‘ecological footprint’ because high bio-capacity has a negative impact on the environment. The variable ‘urbanisation’, which is often associated with job concentration in and around major cities, is expected to have an adverse impact on the environment and is also expected to be positively related to ‘ecological footprint’. As discussed previously, trade openness may have either a positive or negative relationship with the ecological footprint. For a country simply aiming to accelerate growth, trade policies would involve importing energy-intensive technologies, which would raise the ‘ecological footprint’. However, for a country using trade routes to import green technologies to replace existing traditional technologies, the relationship with the ecological footprint is expected to be negative.

In this section, we develop an econometric relationship between ‘ecological footprint’ and different variables, taking into account the ‘N’-shaped EKC hypothesis. We then use co-integration and causality models to determine whether a long-term sustainable relationship exists between the variables.

In this section, we discuss the results of cointegration, causality and model diagnostics, as well as testing the validity of the ‘N’-shaped EKC in the Indian context. For this purpose, we make use of Tables I, II and III, as well as Fig. 2.

Fig. 2. 

CUSUM and CUSUMSQ Plots for the Model Stability

Table 1(a) shows the results of the partial ‘F’ bounds test, which indicate the presence of co-integration between the ecological footprint and the four variables: urbanisation (URB), economic growth (GDP), trade openness (TO) and bio-capacity (BC). This inference was based on a computed ‘F’ value of 13.08, which was much higher than the critical upper bound of 4.9 at a 1% level of significance.

Having proved cointegration in the Indian context, we will now move on to Table I(b), which presents the long-run results obtained from the ARDL model. The results show that, except for trade openness, all the other variables significantly impact the ecological footprint in the long run. The three terms GDP(γ₂), square GDP(γ₆) and cubic GDP(γ₇) in equation (iii), have negative, positive and negative coefficients, respectively, all of which are statistically significant, thus failing to satisfy the ‘N’-shaped EKC curve. Furthermore, the GDP and its transformed coefficients also fail to satisfy the traditional inverse U-shaped EKC curve, which enables us to conclude that the EKC model does not appear to be applicable in the Indian context. The results actually showed an inverse N-shaped curve, which, although not an EKC curve, satisfied the sufficient condition for the existence of real and distinct roots. This could be seen when considering the results obtained under the ARDL long-run model (β₁ = 1432.554, β₂ = –157.394 and β₃ = 5.744923). By substituting these values in the formula (2β₂)² – 4(3β₃)(β₁), we obtain a po sitive value of +332.051.

Moving to the short-run results (Table I(c)), it was observed that the variables ‘urbanization’, ‘trade openness’ and ‘GDP’ impacted ‘ecological footprint’ at their first lags. The results also depicted a negative contemporaneous relationship between ‘ecological footprint’ and the variable ‘biocapacity’. Then, accounting for the sign of the slope coefficients of the variables, we found that the variable ‘urbanization’ was significant with a negative sign, reflecting an attempt towards sustainable urbanization. However, the same variable was positive in the long run, indicating that the country would need to make significant progress before the impact would be visible. The sign of the coefficient of the variable ‘trade openness’ was positive, reflecting the fact that India’s trade had not yet reached a level that could impact ‘ecological footprint’ in a sustainable manner.

Another important result under Table I(c) was provided by the ECM (-1) coefficient, which depicted co-integration equilibrium dynamics with a value of -0.0487483. As this term was both negative and significant, it showed that the movement back towards equilibrium was stable, with a slow adjustment speed covering 4.87% disequilibrium over one period (one year). To investigate causality, we paired the ecological footprint (EF) with each of the explanatory variables: urbanization (URB), economic growth (GDP), trade openness (TO) and bio-capacity (BC). The results presented in Table II reveal the presence of unidirectional causality moving from all four variables to the ecological footprint (EF). The same was also visible in the short-run results.

Finally, Table III shows the results of the model diagnostics tests for serial correlation, heteroscedasticity and model specification, and Figure 2 shows the model stability. All three diagnostics are adequately satisfied, as shown by the results in Table III, i.e. the model does not suffer from serial correlation, is homoscedastic and has no specification error. Further stationary tests, which are also included in Table III, reveal that two of the variables, Ecological Footprint and Biocapacity, are stationary at level, while the rest of the variables, GDP, Urbanization and Trade Openness, are stationary at first difference. The mixed nature of variable stationarity was also the reason for undertaking the ARDL co-integration model.

The present paper examined whether India’s growth–environment trade-off was sustainable. To this end, the study first tested the N-shaped EKC model in the Indian context, followed by the development of a co-integrating relationship for India between an environmental proxy variable (ecological footprint) and a mix of traditional and non-traditional variables, such as GDP, trade openness, urbanisation and bio-capacity. The ARDL model of cointegration was applied, with causality being tested using the Toda-Yamamoto procedure. The two models were considered based on the level of integration among the variables.

The results revealed a co-integrating relationship between the ‘ecological footprint’ and the other variables. These results are consistent with those of several existing studies, including research by Ahmed et al. (2022), Ansari (2020) and Hassan et al. (2019), among others. Furthermore, all the variables except trade openness were found to have a significant impact on the ecological footprint in the long term, with the signs of the coefficients also being as expected. However, neither the N-shaped nor the traditional inverted U-shaped EKC was proven in the study. Therefore, it can be concluded that the EKC model is not applicable to the Indian context. The short-run results showed a lagged relationship between the ecological footprint and GDP, urbanisation and trade openness, while a negative contemporaneous relationship was visible with respect to the variable ‘biocapacity’. The causality results revealed that all variables were moving towards the ‘ecological footprint’. All the model diagnostics, i.e. serial correlation, heteroscedasticity, stability and model specification, were found to be satisfactory.

These study results provide policymakers with useful insights, particularly in countries like India, which has recently made serious attempts to promote sustainable development through their programmes and initiatives. Firstly, with respect to the variable ‘urbanization’, the study found a significant but negative relationship with ‘ecological footprint’ in the short term, and a significant positive relationship in the long term. This suggests that although there is an attempt towards ‘sustainable urbanization’, it may take some time to produce visible results, which largely depends on the state’s commitment and the design of its policies. For example, a viable but challenging measure could be to restrict job opportunities around megacities. However, this requires significant investment to develop the necessary infrastructure in smaller towns, coupled with the design of policies to incentivize firms to open businesses in rural and semi-urban areas.

Another significant relationship which was observed with respect to the variable “Trade Openness”, was that the variable was significantly and positively related to the ecological footprint in the short run, but insignificantly in the long run. This shows that India’s trade is still not sustainable, implying that there is not much discouragement of imports of non-renewable energy products. The state’s role is also important here, as it can encourage more trade in renewable energy through its policies, while gradually shifting towards a complete ban on non-renewable energy. However, this can only be achieved by enacting the country’s medium- to long-term priorities, which may require the postponement or sacrifice of some growth plans in order to achieve the goals of sustainable development.

With regard to India’s GDP, the variable showed a positive relationship with the ‘ecological footprint’ in both the short and long term, thereby providing further evidence that India’s environmental compromises in pursuit of economic growth were still very much in place. However, such a result must be viewed in conjunction with other results, e.g. those obtained under trade openness, to assist in the formulation of precise, targeted policies. Furthermore, as the EKC was not established for India, the study reveals nothing about the ‘magical’ income threshold, which, according to empirical studies, is extremely important for understanding the country’s ability to reverse the negative impact of environmental degradation.

With the Indian government having clearly outlined its roadmap for transforming India into a $5 trillion developed nation by 2047, many schemes have been launched to enable this to be accomplished in a sustained and inclusive manner. These initiatives include the National Solar Mission, the Faster Adoption and Manufacturing of Hybrid and Electric Vehicles (FAME) scheme, the Smart Cities Mission and improved waste management systems, as well as metro rail expansion projects. India also announced at the COP26 Glasgow 2021 summit (UNFCCC, 2021) that it would achieve net-zero carbon emissions by 2070, suggesting that the country’s long-term development trajectory is environmentally sustainable.

Before concluding, we would like to suggest a few areas for future research in this field. Firstly, researchers may wish to explore which other BRICS nations show results similar to those obtained for India in the study, i.e. the presence of a co-integrating relationship but failing on account of validity of EKC Hypothesis. This is also important in light of recent findings by Sun et al. (2026), where research showed that the BRICS countries exhibit altered, inverted EKC N-shaped behaviour. This indicates the need for further empirical evidence on the subject. Another interesting area of research could be the application of nonlinear methodology to modelling similar relations, as the current linear or quasi-linear models with cubic and quadratic terms have been criticized in some studies due to the likelihood of heterogeneity in the findings.

Non-linearity could be introduced into the present model by making NARDL modifications, which would retain the existing structure, or by using quantile methodology, which could additionally provide a comparative assessment of the BRICS countries with respect to environmental concerns. Another suggestion for researchers would be to increase the number of variables. Researchers could consider some uncommon but fine-tuned non-traditional variables, data on some of these variables is now available for research purposes. Such variables include green transport, the quality of governance, corruption and democracy, among others. This would not only have a significant impact on the quality of the research, but also make the research more interesting, meaningful and could easily be extended to cover all the BRICS. Finally, the researchers could consider extending the present sample to include more countries. For example, a good sample could comprise BRICS + countries, or this India-specific study could be extended to include some other Asian countries, as they share many features with India.