Do FDI Flows Lead to Environmental Degradation in Developing Economies? A Case Study of Select Asian Economies

Abstract

Developing countries have an urge to boost economic development and they provide a supportive platform for production units with less stringent environment norms. Hence substantial numbers of manufacturing units are moving from developed economies to developing economies. Such a shift is leading to an upsurge in foreign direct investment (FDI) inflows in the developing economies. The study is an attempt to find relationship between FDI (inflows) and environmental degradation using a sample size of 14 developing Asian economies over the period of 1971–2019 using panel autoregressive distributed lag specifications. Study adopts Environmental Kuznet Curve (EKC) technique to validate the existence of pollution havens in the region. Two panel regression equations were formed: first equation covers the link between economic growth (income per capita) and environment by incorporating square of GDP per capita as one of the explanatory variable. The second equation examines the relation between investment and environmental degradation by accommodating square of FDI as one of the dependent variable. The long-run results for first equation validated the presence of EKC, supporting the presence of pollution havens in the region. Long-run results for non-linear FDI depicted positive and significant outcome (using pooled mean group) indicating negligence towards environment. The results indicated that developing countries are not working towards inviting investment flows which are more environment friendly.

Keywords

FDI Environment EKC Developing Economies Carbon Emissions Panel ARDL

Introduction

Foreign direct investment (FDI) refers to establishing control over the functioning and management of foreign (host) company or setting up production processes on foreign land. Opening gates for foreign investors help the host economy to fulfil large number of benefits such as improvement in balance of payment, access to financial resources, better technology, better employment opportunities and supplement economic growth (Belloumi, 2014). But simultaneously, the host economy also experiences bundle of flaws associated with the establishment of foreign production plants and extension units in terms of increase in stress on existing resources, straining the growth of indigenous industry and environmental degradations (Abdouli & Hammami, 2015; Shahbaz et al., 2015). And in case, the host economy is a developing economy with less stringent FDI norms then the host economy is bound to receive less privileges and more limitations associated with the inflow of foreign investment.

Post globalization, the flow of FDI in developing economies has increased. As per global investment trend monitor (UNCTAD, 2014), the share of developing economies has reached to 52% of total global FDI inflows in 2013. This indicates that developing economies are emerging as most sought after location for FDI (Nunnenkamp, 2001). The present study tries to look into whether this increase in FDI inflows in developing economies is largely due to less rigid (and less protected) environmental norms or not. Developing economies are more inclined towards development rather than sustainability, encouraging the develop economies to shift their production plants in developing economies (Aminu Aliyu, 2005). Such scenario leads to emergence of developing economies as pollution havens wherein production units can easily be established without much of regulatory interventions (Cai et al., 2018). Establishment of more production plants leads to increase in energy demand which in turn leads to more CO₂ emission and hence a negative impact on environment. Moreover, developing economies have lesser economic development, large population (specially unemployed) and untapped natural resources forcing the developing nations to come up with friendly investment policies; offer cheap labour and ease production processes. Therefore, the authorities in developing economies are either working less towards environment (Shahbaz et al., 2011) or they are facilitating production process without taking in account the extent of environmental deterioration leading to emergence of pollution haven hypothesis.

Bowonder (1985) identified two major environment problems that developing countries face: first one is associated with underdevelopment and second is arising due to economic development. Majority of population in developing countries face poverty issues therefore they are highly vulnerable to environmental degradation, hence protection and management of natural resources become a big challenge. Moreover, activities associated with economic development, industrialization and urbanization (Ameen & Mourshed, 2017) further pressurize the environment. Therefore, such scenario creates a mess for developing countries but in order to sustain the needs of huge population, economic development seems to be the only way out and hence sustainability of environment becomes a secondary issue to address.

A significant dichotomy can be seen on world map. Developing economies are struggling to gain growth therefore the major concern is towards development and for them environment norms are taking a back seat Stavropoulos et al., 2018. Whereas the major focus of developed economies is on achieving sustainability along with economic growth hence they are coming up with norms which help to maintain balance between environment and economic growth.

In order to validate presence of environmental degradation, the study employs Environment Kuznets curve (EKC/environmental degradation curve). EKC defines the relationship between environmental degradation variables and income per capita of an economy/region. Kuznets (1955) proposed a relationship between economic development and income inequity by suggesting an inverted U shape curve. He profound that as per capita income increases, income inequality also increases initially but subsequently, it starts falling after a turning point. However, Grossman and Krueger (1991) introduced Kuznet with an inverted U shape/bell shape for establishing relationship between income and environment, Panayotou (1993) named it as Environmental Kuznet curve. EKC is based on the assumption that initially, economies which are at the primary stage/s of development tend to deteriorate environment at the stake of economic development but after reaching to a turning point tend to work more towards environment (sustainability) along with growing economic development. Therefore, EKC (Figure 1) is said to experience an inverted U-shape with environmental degradation and economic development (Grossman & Krueger, 1995; Rothman, 1998; Selden & Song, 1994). And tendency of developing economies to undermine the environmental concerns promotes pollution havens hypothesis.

Figure 1.

Source: Kuznet’s curve.

Trends of FDI Inflows, Economic Growth, Energy Consumption and Carbon Emission

The study tabulates (Table 1) FDI inflows, economic growth, energy consumption and carbon emission for selected developing Asian economies (14 in number) in order to capture the trends of economic growth/investment and environment degradation in Asian region. The current study captures tenure of 49 years (1971–2019), therefore to study the trends 49 years are covered in seven slots. The average annual is calculated in seven slabs and each slab has seven years annual average of all 14 countries captured in the present study. As evident in Table 1, FDI inflows and GDP per capita for Asia’s developing economies have increased in 2012–2019 as compared to inflows in 1971–1977, such trends seem to be satisfactory in terms of economic development for the sample countries.

Table 1.

Trends of FDI Inflows, GDPpc, Energy Consumption and Carbon Emission

Years	GDP Per Capita (US$)	FDI Inflows (US$)	Energy Use Per Capita (kg Oil)	CO₂ Per Capita (Metric Tons)
1971–1977	977.33	78674072.8	832.63	2.26
1978–1984	1738.94	152393732.9	938.6	1.96
1985–1991	1481.96	533473885.1	1008.5	2.21
1992–1998	2076.26	3558936063	1018.29	3.04
1999–2005	3161.58	5722316350	1439.91	3.84
2006–2012	6687.19	19904768512	1686.84	4.51
2013–2019	7250.73	23331478866	1681.25	4.45

Source: Based on data collected from World Bank database.

However, the trends for energy consumption (EN) per capita (in kg oil) and carbon emission per capita (in metric tons) are alarming as increase in both turns up to be more than double since 1971–1977. The trends depict that Asian developing economies are not able to keep a check on environmental degradation. Although, the growth in FDI inflows and GDP per capita is towards well-being but rise in carbon emission along with EN is too huge and definitely needs attention (and remedial measures). Asian countries need to move toward sustainability in order to unleash the best of both economic growth and FDI inflows in long run.

Rationale of the Study

Appraising the trends and literature is essential to assess the gaps in existing studies and to provide a guide to the researcher to work in order to fill the existing gaps.

Following the global trend, Asian economies have also seen an upsurge in the flows of FDI. Moreover, more than half of the Asian economies are still at the growing stage of economic development and trends for carbon emission in the Asian region indicate negligence towards environment. Our present study is an attempt to empirically examine the impact of FDI inflows on environment in developing Asian economies. Using EKC, the study also tries to examine whether the developing economies in the region will witness a rise in pollution level with increase in economic growth (income per capita) and/or FDI inflows. In order words, the study tries to examine whether pollution haven hypothesis is true for the developing Asian economies or not.

Moreover, an insight into the existing literature exhibits that a number of studies have captured the impact of economic growth and investment on environment for Asian region but most of them are country specific rather than region specific. Alam (2014) captured the impact on environment for Bangladesh; Hitam and Borhan (2012) studied the EKC analysis for Malaysia. Tiwari et al. (2013), Gupta and Yadav (2014) and U. and Mitra (2020) examined EKC hypothesis for India. Jayanthakumaran et al. (2012) studied trade, growth and energy for India and China. Ahmed and Long (2012) investigated EKC for Pakistan. Only a couple of studies (Gunarto, 2020; To et al., 2019) have examined the impact of economic growth and investment flows on environmental degradation for emerging Asian economies collectively. Our study is also an attempt to empirically examine the impact of economic development and FDI inflows on environment for select Asian developing economies using two separate regression equations (one for economic growth and other for FDI flows) unlike the existing studies (To et al., 2019) that have examined both economic growth and investment simultaneously as part of a single regression equation.

Literature Review

Numbers of studies have presented and empirically examined the nexus between environmental degradation– investment flows, and environmental degradation– economic growth using varied sample size, regional coverage and empirical tools. Few studies have country specific coverage. Kheder (2010) tried to empirically examine the relationship between FDI, environmental regulation and pollution using FDI flows for the period 1999–2003 by employing three simultaneous equations. The study found a negative impact of environmental regulation on FDI and French manufacturing FDI was found to increase the pollution emissions in host countries. Balaguer and Cantavella (2016) conducted structural analysis for Spain using EKC for the period 1874–2011. The study found that EKC existed in the region, both for long run as well as short run. Similarly, Dasgupta et al. (2002) also conducted the study for Spain, but the results were not supportive. Boluk and Mert (2015) conducted the study for Turkey using EKC framework and adopting autoregressive distributed lag (ARDL) approach for period 1961–2010. The study found the existence of inverted U-shaped for per capita emissions and per capita real income both in long and short run. Tutulmaz (2015) also examined the EKC analysis between CO₂ emissions and GDP per capita for Turkey using time series analysis for the period 1968–2007.

Alam (2014) examined changes in economic structure and trends of CO₂emissions with GDP per capita of Bangladesh for the period 1972–2010. The link between economic growth (GDP per capita) and CO₂ emissions of Bangladesh was examined using EKC hypothesis. The results indicated that existence of EKC inverted ‘U’ shaped does not hold true for Bangladesh. Liang (2006) also captured significance of EKC curve for China. The study captured the validity of EKC using sulphur dioxide as the dependent variable. The results suggested a negative correlation between FDI and air pollution, suggesting that the overall effect of FDI may be beneficial to the environment. Li et al. (2016) examined EKC hypothesis for 28 Chinese provinces using generalised method of moments (GMM) and ARDL framesets whereas Wang et al. (2015) captured 30 Chinese provinces for the sample period 2001–2010.

Gupta and Yadav (2014) tried to examine the relationship between CO₂ emissions and some of macroeconomic factors responsible for change in CO₂ emissions. The study tried to explore whether there exists an inverted U shape via establishing relationship between per capita GDP and CO₂ emissions as hypothesized by EKC using panel data for the period 1981–2006. The study revealed that India has not yet reached to the turning point so far. Tiwari et al. (2013) also examined EKC hypothesis for Indian economy by employing ARDL model for the sample period 1966–2009. The study supported the presence of EKC hypothesis both in long and short run.

Few studies have also captured results for more than one developing country simultaneously. Choi et al. (2010) examined the existence of the EKC for CO₂ emissions and its causal relationships with economic growth and openness by using time series data for the period 1971–2006 from China (an emerging market), Korea (a newly industrialized country), and Japan (a developed country). The result for the study depicted large heterogeneity among the sample countries. Jayanthakumaran et al. (2012) examined EKC framework for India and China using ARDL methodology. Sharma et al. (2020) examined whether pollution level in five South Asian countries has enhanced as a result of trade liberalization and financial crisis or not by employing GMM for the period 1981–2015. Though the results for trade liberalization predicted increase in carbon emission in the region, the results for financial crisis for long run were found to be insignificant. Rahman et al. (2020) captured EKC for Bangladesh, China, India and Myanmar using ARDL model. The results indicated that EKC existed for India and China both in long as well as short run whereas, for Bangladesh and Myanmar EKC framework was true only for short run.

Studies incorporating large sample size with number of countries have also been identified. Kahouli (2014) employed a sample size of 14 home countries and 39 host countries for the period 1990–2011 in order to capture the impact of environment by employing static and dynamic gravity equations and simultaneous equations. The study found varied results for trade but the result for FDI is found to be positive but insignificant. Shahbaz et al. (2011) conducted a study to empirically examine the nexuses between economic growth–environment and foreign investment–environment using sample of 110 developing and developed countries by applying a pooled regression, fixed effects and random effects model. The study validated the presence of inverted U-shaped EKC for the said sample. Hill and Magnani (2002) examined the EKC framework for a panel of 156 countries by employing generalized least squares model. But the study found no evidence of an inverted U-shaped EKC hypothesis. Shahbaz et al. (2015) tried to examine non-linear relationship between FDI and environmental degradation for high, middle and low-income countries in a multivariate framework using economic growth and EN. Long-run results estimated by applying fully modified ordinary least squares (FMOLS) suggest that EKC exists and FDI increases environmental degradation. Ng et al. (2020) examined EKC for 76 countries using common correlated effects of mean group estimation. The study suggested that only 16 out of 76 countries supported EKC presence.

Some studies have even examined environmental degradation for blocs/group of similar countries. Neagu (2019) examined EKC framework for 25 EU countries by employing cointegration polynomial regression both for panel as well as time series framework. The results supported presence of economic complexity and carbon emissions for EU countries. Aminu Aliyu (2005) found that FDI outflow is positively correlated with environmental policy in 11 organisation for economic co-operation and development (OECD) countries. However, FDI inflow is not significant in capturing the level of pollution and energy use for non-OECD countries. Zakarya et al. (2015) studied Brazil, Russia, India, China and South Africa (BRICS) countries for the interactions between EN, FDI, economic growth and CO₂ emissions for a sample period of 1990–2012. Abdouli and Hammami (2017) examined the relation between economic growth, FDI inflows and EN using GMM specification for a sample of 17 from Middle East (ME) and North Africa (NA). The results indicated a bidirectional causal relationship between FDI inflows and economic growth, EN and economic growth and unidirectional relationship between EN and FDI inflows. Similarly, Shahbaz et al. (2019) captured the relationship between FDI and carbon emissions for MENA region by employing GMM specifications and the results suggested presence of pollution haven in the region.

Running down the line, we were able to identify rich literature capturing economic growth, investment flows and environment degradation using EKC hypothesis, hence inducing us to examine the presence of EKC framework in select Asian economies specifically for post-liberalization.

Research Methodology

Sample Size and Data Source

Massive industrialization, liberalization and globalization contributed to global warming and industrial pollution. In order to extensively capture the impact of industrialization on Asian developing economies, the study captures a long tenure of 49 years starting from 1971 till 2019 to study the impact of FDI inflows (and economic growth) on environmental degradation. The study captures 14 developing (developing as per World Economic Situation and Prospects 2012) Asian economies to examine the impact on environment. The countries covered were China, Bangladesh, India, Sri Lanka, Turkey, Jordan, Thailand, Nepal, Indonesia, Pakistan, Kuwait, Maldives, Malaysia and Philippines. In 2014, these countries received 18.89% of the total global FDI net inflows whereas these countries received around 29% of total global FDI in 2018 (Table 2). Some countries such as United Arab Emirates, Iran, Myanmar, Cambodia and Afghanistan were dropped due to non-availability of data for explanatory variables.

In order to study the impact of foreign investment on environment, the data for GDP per capita, CO₂ emissions (metric tons per capita), FDI net inflows (in current US$) and energy use (kg of oil equivalent per capita) was gathered from World Bank database.

Table 2.

Foreign Direct Investments, Net Inflows for 14 Developing Asian Economies

Country Name	2014 (US$) million	2018 (US$) million
China	268097.00	235365.00
Bangladesh	2539.19	2421.63
Indonesia	25120.73	18909.83
India	34576.64	42117.45
Pakistan	1887.00	1737.00
Philippines	5739.57	9948.60
Jordan	2178.45	954.93
Sri Lanka	893.63	1614.04
Turkey	13337.00	13023.00
Thailand	4975.46	13205.11
Maldives	333.38	575.66
Malaysia	8570.09	10619.43
Nepal	30.40	68.19
Kuwait	485.81	–21.38
Total FDI inflows for 14 Asian developing countries	368765.00	350539.00
World	1952540.00	1189220.00
% Share of world FDI inflows	18.89	29.48

Source: Data collected from World Bank database.

Methodology

In order to empirically examine the impact of FDI and economic growth on environment for 14 Asian countries over the period of 49 years, the study adopts following procedure:

Panel unit root tests: In order to estimate reliable relationship between variables, we need to check the stationarity properties of the data under consideration. Variables that are non-stationary generate results that are spurious and unreliable. In order to check the stationarity, panel unit root testing is recommended.

Due to two-dimensional nature of the data and to check the non-stationarity of the data, the study applies panel unit root testing suggested by Im et al. (2003) and Fisher-Augmented-Fuller test (ADF) and PP in order to accommodate the variations for FDI, EN, GDP per capita and CO₂.

Panel co-integration test: In order to examine the presence of a long-run relationship between FDI and CO₂, and economic development and CO₂, the study uses a Fisher-type test using an underlying Johansen methodology (Maddala & Wu, 1999) and the Pedroni residual cointegration test (Pedroni, 1999, 2004). Pedroni residual cointegration test (panel cointegration test) accounts for heterogeneous intercept and also takes care of trend coefficient across the cross entities (countries). Pedroni residual cointegration provides two types of statistics, namely, panel statistic and group-mean statistics. First-order autoregressive term for panel statistics is same across all the countries and for group statistics, it varies over the countries.

Panel autoregressive distributed lag (ARDL) model:

In order to examine the long-run coefficients, the study employs panel ARDL model. ARDL is ornamented with various advantages unlike other long-run dynamic methods (such as FMOLS, Dynamic Ordinary Least Square and GMM). Firstly, panel ARDL model takes lag terms of both dependent as well as independent variables, helping the researcher to fix the issues associated with endogeneity. Secondly, ARDL is efficient enough to compute both long-run and short-run estimates in one go for a data specification having both cross-sectional and time dimensions aspects (Samargandi et al., 2013; Ullah & Awan, 2019). Lastly, the technique is well suitable for data where the variables under study are either integrated at level, first difference or a combination of different order exists (Pesaran & Shin, 1999).

Panel ARDL employs three model estimations namely, pooled mean group (PMG) model suggested by Pesaran et al. (1999), mean group (MG) specifications developed by Pesaran and Smith (1995) and dynamic fixed effects (DFE) model. All three specifications are based on maximum likelihood estimations and PMG lies at a point in middle of MG and DFE, accounting for averages and pooling of coefficients (Blackburne & Frank, 2007). The choice of model (PMG or MG or DFE) for result interpretation depends upon Hausman testing. For selection between PMG and MG specifications, the null hypothesis for Hausman test supports use of coefficient generated via MG model as more appropriate and alternative hypothesis suggests use of PMG model. Similarly, Hausman test also helps the researcher to select between PMG and DFE. PMG specification for ARDL is generally formulated as:

y_{i t} = \sum_{j = 1}^{p} γ_{i j} y_{t - j} + \sum_{j = 0}^{q} β_{i j} x_{t - j} + μ_{i} + ε_{i t}

where i = 1, 2, …, n countries and t = 1, 2, …, t number of years,

y_it denotes dependent variable (CO₂),

x_t _{− j} denotes explanatory variables such as FDI, FDI square, GDP per capita, GDP

per capita square and energy usage,

ε_it represents error term, and

µ_i represents the group-specific effect.

Causality test: For identifying the direction of relationship between the variables, the study adopts Dumitrescu and Hurlin (DH) causality tests for panel data.

The study tries to validate whether Environmental Kuznets curve exists (inverted U-shape curve) and presence of pollution havens hypothesis can be seen in the Asian subcontinent.

Model Specification

The study is an attempt to capture the impact of FDI, economic growth and EN on CO₂ (environment degradation variable) therefore the basic model specification formed to capture the above said variable is as follows:

C O_{2 i t} = β_{1} + β_{2} F D I_{i t} + β_{3} Y p c_{i t} + β_{4} E N_{i t} + ε_{i t}

(1)

However, the study employs double log model for the basic panel regression to analyse the presence of EKC curve, therefore the model specification for the current study can be written as:

\begin{matrix} \ln C O_{2 i t} = β_{1} + β_{2} \ln F D I_{i t} + β_{3} \ln Y p c_{i t} \\ + β_{4} \ln E N_{i t} + ε_{i t} \end{matrix}

(2)

where βs are regression coefficient

CO_2it represents carbon dioxide emissions (metric tons per capita) for country i,

FDI_it represents foreign direct investment net inflows (in current US$) for country i,

EN_it represents energy use (kg of oil equivalent per capita) for country i,

Ypc_it represents income (GDP) per capita for country i, ε_it represents error term.

Explanation of the Variables

Carbon emission (CO₂ ) has been incorporated to measure the level of environmental degradation. Combustion of fossil fuel is source of carbon emission and this leads to deterioration of environment. Combustion of fossil fuel is generally due to electricity generation, transportation, industrial processes and energy generation, and also changes in land uses leads to combustion. Carbon emission is leading to global warming, hence degrading environment. Shahbaz et al. (2015), Omri et al. (2014), Alam (2014) and Beak and Koo (2009) examined environmental degradation using CO₂ as dependent variable.

Foreign direct investment net inflow (FDI) has been included to examine whether inflows of foreign investment in Asian economies are leading to deterioration of indigenous environment or not. Seker et al. (2015) and Shahbaz et al. (2015) examined the impact of FDI on environment. With increase in the inward flow of FDI, the number of production processes increases leading to a rise in emissions. However, in case the host economy has less stringent environment norms then the firms are not forced to set up plants with pollution controllers or set up plants which are environment friendly then a rise in FDI inflows may have a negative impact on environmental degradation. Hence, economies with less rigid environment norms (pollution havens hypothesis) will experience positive relationship between FDI inflows and environmental degradation (CO₂) whereas countries with stringent environment standards will witness negative relation between FDI inflows and environmental degradation.

Energy consumption (EN) has been encompassed to examine the impact of EN on environment. Larger is the consumption of energy in an economy, more is the impact on environment degradation (CO₂). Tamazian and Bhaskara (2010), Linh and Lin (2012), Shahbaz et al. (2015) and Khan et al. (2014) have incorporated EN as one of the explanatory variable for environmental degradation. We expect the variable to have positive relation with environmental degradation.

Income (GDP) per capita has been employed to examine the impact of the level of economic development on environment degradation. Countries which are at the earlier stages of development may forgo environment norms as economic growth (in terms of industrialization) seems to be more significant. Hence developing economies will have positive relationship between GDP per capita and environment degradation. Moreover, growing economies tend to promote industrialization therefore they are bound to have more emissions. Whereas economies that are more developed tend to have learning effects and looks for sustainability, hence with economic growth, developed economies try to come up with establishments which are either less polluting or tend to reduce pollution. Hence, a negative relationship between economic growth and environmental degradation can be seen for a developed and environment vigilant economy.

In order to study (validate) the presence of Environmental Kuznets curve, we form two linear relationships (two models): model one between economic growth and environmental degradation (Equation 3) and second model between FDI and environmental degradation (Equation 4). The modified form of basic equation (Equation 2) which will take care of Kuznets curve for income (GDP per capita) and investment can be represented as follows:

\begin{matrix} \ln C O_{2 i t} = α_{1} + α_{2} \ln F D I_{i t} + α_{3} \ln Y p c_{i t} \\ + α_{4} {(\ln Y p c_{i t})}^{2} + α_{5} \ln E N_{i t} + ε_{t} \end{matrix}

(3)

\begin{matrix} \ln C O_{2 i t} = β_{1} + β_{2} \ln F D I_{i t} + β_{3} \ln Y p c_{i t} \\ + β_{4} {(\ln F D I_{i t})}^{2} + β_{5} \ln E N_{i t} + ε_{t} \end{matrix}

(4)

Equation 3 has been incorporated to validate the existence of EKC curve, that is, inverted U-shape relationship between environmental degradation and economic growth. GDP per capita and square of GDP per capita have been incorporated to examine the impact of development stages. For Equation 3, α₃ (coefficient for gross domestic product per capita [GDPpc]) is bound to be > 0 and α₄ (coefficient for square of GDPpc) will generate a value < 0 in case the EKC curve existence in the Asian region is experienced. Liang (2006) and Choi et al. (2010) have examined the relationship between economic development and environmental degradation.

Similarly, Equation 4 examines the relationship between environmental degradation and FDI inflows. Studies (Shahbaz et al., 2015) capturing the impact of FDI on environment have examined the impact of FDI and square of FDI on the environment deterioration to validate the presence of EKC curve for investment. Again, β₂ is bound to be > 0 and β₄ will generate a value < 0 in case the EKC curve exists in the Asian region and pollution haven hypothesis exists for investment flows. Therefore, Equation 3 studies FDI as one of the explanatory variable in the regression capturing economic development via EKC curve whereas Equation 4 exclusively covers the relation between investment inflows and environmental degradation.

Results and Analysis

Results for Panel Unit Root Testing

As discussed earlier, the study tries to examine the presence of stationarity for FDI, EN, GDP per capita/economic growth and carbon emissions using panel unit root testing. Im et al. (2003) and Fisher-ADF and Phillips & Perron (PP) tests were performed at level and first difference (Table 3). We may conclude from the results (using all model specifications) that GDP per capita, carbon emissions, EN and FDI are stationary either at first difference or at level. The unit root testing result indicates mixed order stationarity, therefore it is necessary to find whether long-run cointegration exists among the variables or not.

Table 3.

Panel Unit Root Test Results

Variable	At Level			First Difference
Variable	IPS W-stat^#	ADF Fisher Chi Square	PP Fisher Chi-Square	IPS W-stat	ADF Fisher Chi Square	PP Fisher Chi-Square
lnCO₂	0.42367	33.6795	36.4958	−13.3831*	223.844*	392.942*
	(.6641)	(.2116)	(.1304)	(.000)	(.000)	(.000)
lnFDI	−3.38026*	53.859*	123.087*	−22.0794*	361.91*	431.782*
	(.0004)	(.0023)	(.000)	(.000)	(.000)	(.000)
lnGDPPC	2.60201	16.7862	18.0126	−13.9008	232.587*	338.302*
	(.9954)	(.9527)	(.9258)	(.000)	(.000)	(.000)
lnEN	1.1565	34.4354	33.8495	−12.0237*	198.041*	373.36*
	(.8763)	(.1869)	(.2059)	(.000)	(.000)	(.000)

Source: The author (calculated by using eviews).

Notes: Null hypothesis: There is unit root.

p-values in parentheses.

*1% level of significance.

^#IPS W-stat indicates Im, Pesaran and Shin W-stat.

Cointegration Test Results

Cointegration results were generated using Fisher panel cointegration and Pedroni residual cointegration test. The results for Pedroni residual cointegration test (depicted in Table 4) indicate that we can reject the null hypothesis (no cointegration) in most of the cases (4 out of 7 statistics) hence presence of cointegration for FDI and economic growth and carbon emission can be seen for panel data for the period 1971 till 2019. The results for fisher, shown in Table 5 panel cointegration supported the results. The null hypothesis for Fisher stat from trace test and from max-Eigen test is rejected at 1% significant level. The results for almost 1 Fisher stat from trace test and from max-Eigen test also reject the null hypothesis for no cointegration at 1% significance level.

Table 4.

Pedroni Residual Cointegration Test Results

	Panel Cointegration Statistic	Group-mean Panel Cointegration Statistic
Variance-statistic	0.397911 (0.3453)
Rho-statistic	−0.144849 (0.4424)	−0.06863 (0.4726)
PP statistic	−2.9249* (0.0017)	−6.27698* (0.000)
ADF statistic	−3.62355* (0.0007)	−7.5512** (0.0469)

Source: The author (calculated by using eviews).

Note: Null hypothesis: No cointegration.

* significant at 1%

**Significant at 5%

Table 5.

Johansen Fisher Panel Cointegration Test Result

Hypothesized No. of Cointegrating Equations	Fisher Stat (From Trace Test)	Fisher Stat (From Max-Eigen Test)
None	359.6* (0.0000)	241.4* (0.0000)
At most 1	163.9* (0.0000)	93.45* (0.0000)

Source: The author (calculated by using eviews).

Notes: p-values in parentheses.

*Significant at 1%.

Results for Panel ARDL

As discussed earlier, the study adopts panel ARDL specification to estimate long-run coefficients. Moreover, the results for stationarity (having mixed order) support employing panel ARDL specifications. The lag structure for ARDL (1,1,1,1,1) was based upon least value of Hannan–Quinn information criterion for Model 1 (Table 6) and Schwartz Bayesian Information Criterion for Model 2 (Table 7).

Table 6.

Panel Vector Autoregression Model for Lag Order Selection (for Model 1)

Lag	LogL	LR	FPE	AIC	SIC	HQ
0	1058.714	NA	3.83e−07	−3.424395	−3.395672	−3.413227
1	1175.289	231.2581	2.76e−07	−3.750938	−3.607326*	−3.695099*
2	1185.266	19.66182	2.82e−07	−3.731382	−3.472880	−3.630871
3	1202.858	34.44234	2.80e−07	−3.736552	−3.363161	−3.591370
4	1237.306	66.99456*	2.64e−07*	−3.796448*	−3.308167	−3.606594

Source: The author (calculated by using eviews).

Notes: Endogenous variables: D(LNEN) D(LNFDI) D(LNGDPPC2) D(LNGDPPC);

Exogenous variables: C.

LR indicates sequential modified LR test statistic (each test at 5% level).

AIC: Akaike information criterion; FPE: Final prediction error; HQ: Hannan–Quinn information criterion; SIC: Schwarz information criterion.

*Lag order selected by the criterion.

Table 7.

Panel Vector Autoregression Model for Lag Order Selection (For Model 2)

Lag	LogL	LR	FPE	AIC	SIC	HQ
0	−1386.273	NA	0.001073	4.513875	4.542597	4.525042
1	−1231.399	307.2356	0.000683	4.062982	4.206594*	4.118821*
2	−1221.866	18.78708	0.000698	4.083980	4.342481	4.184490
3	−1199.885	43.03385	0.000684	4.064561	4.437952	4.209743
4	−1169.721	58.66198*	0.000654*	4.018576*	4.506857	4.208429

Source: The author (calculated by using eviews).

Notes: Endogenous variables: D(LNEN) D(LNFDI) D(LNFDI2) D(LNGDPPC);

Exogenous variables: C.

LR indicates sequential modified LR test statistic (each test at 5% level).

AIC: Akaike information criterion; FPE: Final prediction error; HQ: Hannan–Quinn information criterion; SIC: Schwarz information criterion.

*Lag order selected by the criterion.

Though the coefficients for Model 1 (Equation 3) are captured using PMG, MG and DFE, but PMG results are used for interpretation. In order to choose between PMG and MG, Hausman test was employed and the p-value rejected the null hypothesis hence opting for PMG for result interpretation. Similarly, the Hausman results for PMG and DFE rejected the null hypothesis (Table 8) and accepted the alternative hypothesis, suggesting PMG for outcome analysis. Therefore, in order to examine the presence of EKC curve in Asian developing countries, PMG was found to be apt. The long-run results (employing PMG) for GDP per capita and square of GDP per capita indicated that an inverted U-shaped (for economic growth) existed in the region. Hailemariam et al. (2019) and Ullah and Awan (2019) also supported presence of EKC curve for developing Asian economies. Similarly, the coefficients for FDI and EN bear a positive and significant result. The results indicate that the growth of foreign investment and EN accompanied by economic growth is leading to a rise in carbon emissions (environmental degradation). The results for EN indicate a positive relationship with carbon emissions, an increased use of energy in terms of oil usage is leading to environment degradation. Developing economies in the region are definitely having weaker environment norms or forgoing a check on environment for the need of development. The results for GDP per capita and square of GDP per capita validate the presence of EKC curve in the region of Asia, indicating a positive relation between income per capita and CO₂ at initial level of development and negative relation after a turning point with rise in income per capita. However, the short coefficients for GDP per capita and square of GDP per capita were found to be insignificant but with relevant signs.

Table 8.

Result for Panel⁴ ARDL³ (PMG/MG/DFE)—Model 1

	Panel PMG-(a)			Panel MG-(b)			Panel DFE-(c)
	Coef.	Std Err	P-value	Coef.	Std Err	P-value	Coef.	Std Err	P-value
Long-run Coefficients				Dependent Variable ΔlnCO₂
lnGDPpc	0.422*	0.088	.000	25.411	24.181	.293	1.032***	0.582	.076
lnGDPpc²	−0.067*	0.011	.000	−3.099	2.911	.287	−0.128	0.092	.164
lnEN	1.226*	0.050	.000	−3.030	4.983	.543	0.909*	0.200	.000
lnFDI	0.009*	0.002	.000	−0.206	0.168	.219	−0.014	0.019	.478
Short-run Coefficients
Er.Corr.	0.243*	0.066	.000	0.324*	0.056	.000	0.075*	0.020	.000
ΔlnGDPpc	0.778	1.018	.445	0.988	1.180	.402	0.193	0.165	.241
ΔlnGDPpc²	−0.181	0.209	.386	−0.222	0.253	.380	0.001	0.025	.974
ΔlnEN	0.934*	0.142	.000	1.043*	0.163	.000	0.289*	0.032	.000
ΔlnFDI	0.000	0.001	.955	0.001	0.002	.677	0.000	0.001	.936
Constt	0.99*	0.268	.000	1.215**	0.569	.033	0.328*	0.103	.001
Hausman test²		(a) vs (b)		chi2(4) = 4.53		(a) vs (c)		Chi2(4) = 5.27
				Prob. > chi² = 0.3394				Prob. > chi² = 0.2611

Source: The author (calculated by using stata¹).

Notes: ARDL: Autoregressive distributed lag; CO₂: Carbon dioxide; DFE: Dynamic fixed effects; EN: Energy consumption; FDI: Foreign direct investment; MG: Mean group; PMG: Pooled mean group.

*1% significance level.

** 5% significance level.

***10% significance level.

¹Estimations done by employing (xtpmg) in Stata.

²Hausman test indicates use of PMG over MG and DFE.

³The lag structure employed is ARDL (1,1,1,1,1).

⁴Annual data for 14 countries 1971–2019.

The study also tries to capture the relation between environment and FDI by squaring the FDI inflows in a separate equation (Model 2). Though the Hausman results used for choosing between PMG and MG reject the null hypothesis, the results for Hausman test evaluating between PMG and DFE accepted the null hypothesis therefore the results for Model 2 are studied for both PMG and DFE models ((Table 9). For PMG specifications, the long-run result for investment does not indicate existence of Environmental Kuznets curve (inverted U) for investment and carbon emission. Result for both linear as well as non-linear flows for FDI is found to be positive, this indicates Asian developing economies are not able to witness a turning point. Results for Equation 4 suggest lack of pollution check in the region for sake of increasing investment flows. Shahbaz et al. (2015) captured similar results for less-developing economies. The long-run and short-run results for DFE for investment were found to be insignificant. The long run as well as short-run results for EN and GDP per capita (both PMG and DFE for Model 2) were found to be positive and significant.

Table 9.

Result for Panel⁴ ARDL³ (PMG/MG/DFE)—Model 2

	Panel PMG-(a)			Panel MG-(b)			Panel DFE-(c)
	Coef.	Std Err	P-value	Coef.	Std Err	P-value	Coef.	Std Err	P-value
Long-run Coefficients			Dependent Variable ΔlnCO₂
lnGDPpc	0.172*	0.029	.000	4.036	3.787	.287	0.337***	0.195	.083
lnFDI	0.111**	0.021	.000	0.066	0.149	.658	0.033	0.061	.584
lnFDI²	0.009*	0.001	.000	−0.093	0.087	.284	−0.005	0.007	.428
lnEN	1.325***	0.064	.000	2.708	1.318	.040	0.894*	0.231	.000
Short-run Coefficients
Er. Corr.	0.115*	0.046	.012	0.403*	0.099	.000	0.066*	0.018	.000
ΔlGDPpc	0.011	0.067	.870	0.041	0.059	.485	0.201*	0.038	.000
ΔlnFDI	−0.016	0.017	.346	−0.008	0.023	.716	0.004	0.005	.370
ΔlnFDI²	0.002	0.001	.174	0.001	0.001	.529	0.000	0.001	.389
ΔlnEN	1.031*	0.155	.000	1.144*	0.207	.000	0.285*	0.031	.000
Constt	0.327*	0.127	.010	1.674*	0.617	.007	0.23*	0.051	.000
Hausman test²		(a) vs (b)		chi²(4) = 4.48		(a) vs ( c)		chi²(4) = 11.12
				Prob. > chi² = 0.3452				Prob. > chi² = 0.0252

Source: The author (calculated by using stata¹).

Notes: ARDL: Autoregressive distributed lag; CO₂: Carbon dioxide; DFE: Dynamic fixed effects; EN: Energy consumption; FDI: Foreign direct investment; MG: Mean group; PMG: Pooled mean group.

*1% significance level.

**5% significance level.

***10% significance level.

¹Estimations done by employing (xtpmg) in Stata.

²Hausman test indicates use of PMG and DFE over MG.

³The lag structure employed is ARDL (1,1,1,1,1).

⁴Annual data for 14 countries 1971–2019.

The overall long-run results (for both panel regression models) indicate that pollution haven hypothesis exists for income per capita regression/Model 1, whereas for investment inflows (Model 2) it was seen that increase in FDI inflows elevates environment concerns for emerging Asian economies.

Results for Causality

DH panel causality tests was performed on the panel data. The results for causality test are presented in Table 10. In order to understand the direction of causality, pictorial presentation is also depicted in Figure 2. Results for EN and carbon emission are found to be unilateral but the flow indicates EN is leading to carbon emission in the region. The result indicated bi-directional flow for FDI and carbon emission, FDI² and carbon emission, FDI and EN and FDI² and EN. The outcome for FDI, carbon emission and EN indicates that the increased use of FDI is enhancing the use of EN and that in turn is increasing the carbon emission. Two way results were also depicted for income per capita (GDPPC) and carbon emission and GDPPC² and carbon emission. However, the results for EN and GDPPC and EN and GDPPC² were found to be unidirectional.

Table 10.

Pairwise Dumitrescu and Hurlin Panel Causality Tests (Sample 1971–2018)

Causality Direction	W-Stat.	Zbar-Stat.	Prob.	Causality Direction	W-Stat.	Zbar-Stat.	Prob.
EN→CO₂	8.30755^*	9.92652	0.0000	GDPPCEN	3.16056	1.54675	0.1219
CO₂→EN	2.25929	0.0794	0.9367	ENGDPPC	6.96384^*	7.73883	0.0000
FDICO→₂	7.20669^*	8.13421	0.0000	GDPPC²EN	2.54791	0.54928	0.5828
CO₂→FDI	8.40569^*	10.0863	0.0000	ENGDPPC²	6.13543^*	6.3901	0.0000
FDI²→CO₂	6.74424^*	7.3813	0.0000	FDI²FDI	8.63649^*	10.4621	0.0000
CO₂→FDI²	10.8651^*	14.0905	0.0000	FDIFDI²	8.82924^*	10.7759	0.0000
GDPPC→CO₂	6.24653^*	6.57098	0.0000	GDPPCFDI	9.7094^*	12.2089	0.0000
CO₂→GDPPC	5.59298^*	5.50695	0.0000	FDIGDPPC	5.91296^*	6.0279	0.0000
GDPPC²→CO₂	5.60012^*	5.51857	0.0000	GDPPC²FDI	8.33581^*	9.97253	0.0000
CO₂→GDPPC²	5.72192^*	5.71687	0.0000	FDIGDPPC²	7.55429^*	8.70015	0.0000
FDI→EN	3.96957^*	2.86389	0.0042	GDPPCFDI²	10.551^*	13.5791	0.0000
EN→FDI	7.21541^**	8.14841	0.0000	FDI²GDPPC	5.41533^*	5.21772	0.0000
FDI²→EN	3.53503^*	2.15642	0.0311	GDPPC2FDI²	11.1696^*	14.5863	0.0000
EN→FDI²	6.89186^*	7.62165	0.0000	FDI²GDPPC²	9.32653^*	11.5855	0.0000
GDPPC²→GDPPC	5.50926^*	5.37064	0.0000
GDPPC→GDPPC²	5.97966^*	6.1365	0.0000

Source: The author (calculated by using eviews).

Notes: EN: Energy consumption; FDI: Foreign direct investment.

*1% of significance level.

**5% of significance level.

Figure 2.

Source: The Authors. Deduced from Table 10.

Conclusion and Policy Recommendations

The study was an attempt to examine the relationship between FDI, economic growth, EN and CO₂ emissions using data for 14 developing countries from Asian. The study applied panel root testing, panel cointegration and causality analysis techniques for the period of 1971–2019. Stationarity test suggested mixed order at level and for first difference. Finally, panel ARDL was used in order to validate the presence of Environmental Kuznets curve, two separate equations (models) were examined, one by squaring GDP per capita to capture relationship between pollution and economic growth (with FDI inflows as one of the explanatory variable) and other by squaring FDI inflows to estimated link between pollution and investment. The model for income per capita supported the existence of inverted U-shape EKC curve in the region. The study found pollution havens in the region; might be due to out-dated technology, less stringent of environment norms, development a priority, industrialization at its primary stage, and lack of environment friendly (alternative) techniques of production. Similarly, the results for FDI linear and FDI non-linear flows suggested environmental degradation, might be due to less concerns for global warming and climate changes. The results indicated that Asian developing economies are more concerned with growth and industrialization rather than environment and sustainability. Might be currently, the developing economies are forced to give more attention to food security, industrialization, employment and urbanization rather than environment.

Though, some significant aspects of environment degradation for Asian developing economies were identified but certain limitations of the current study can also be stated. Firstly, the Asian countries selected were only 14 (in size), largely due to non-availability of data for some other developing countries (Myanmar, Cambodia, Afghanistan, etc.) of Asia. Other limitation of the study was that we were not able to identify the nature/type/sector of firms which are major pollution contributors in Asian region as firm level data for FDI and CO₂ was not accessible.

However, the developing countries should sincerely realize that steady economic development without environmental protection will reap shallow results of development in long run; the impact of development will be less felt or may even negate in case the nature is ignored.

The use of sustainable techniques of production, namely, biomass, renewable energy, fitting of pollution controllable set-up and geothermal heat and sunlight will surely help them to reduce CO₂ emission and also support growth. The developing economies should initiate sustainable measures so that more consistent and durable (in terms of environment) growth can be seen.

Moreover, sincere and steady efforts are required to save and protect environment. Developing countries should contribute towards R&D so that they come up with pollution control units or alterative (environment friendly) production units.

Even, emerging economies should strengthen legislations for environment protection, provide a check mechanism on polluting units (vehicles, plants and other devices) both for indigenous and foreign corporates, monitor (and also identify) pollution generation industry wise and product wise, impose penalties on polluting plants such as carbon tax, encourage pollution reducing units by monetary and non-monetary initiatives such as tax relaxation and provide incentives to plants working towards recycling of waste. Moreover, electronic products should essentially carry energy conservation marks and also bear details of the energy conservation done by company to which the product belongs.

Developing countries should place strong norms (taking care of environment) especially while negotiating treaties and agreements supporting FDI inflows.

A concern only towards industrialization may yield momentary gains but will have a negative and deteriorating impact in the long run. Without a healthy environment and healthy citizens, massive industrialization (growth) will be of no use.

Footnotes

Declaration of Conflicting Interests

The author declared no potential conflicts of interest with respect to the research, authorship and/or publication of this article.

Funding

The author received no financial support for the research, authorship and/or publication of this article.

References

Abdouli

, & Hammami

(2015). The impact of FDI inflows and environmental quality on economic growth: An empirical study for MENA countries. Journal Knowledge of Economics. https://doi.org/10.1007/s13132-015-0323-y.

Abdouli

, & Hammami

(2017). Economic growth, FDI inflows and their impact on the environment: An empirical study for the MENA countries. Quality & Quantity, 51, 121–146. https://doi.org/10.1007/s11135-015-0298-6

Ahmed

, & Long

(2012). Environmental Kuznets Curve and Pakistan: An empirical analysis. Procedia Economics and Finance, 1(1), 4–13.

Alam

(2014). On the relationship between economic growth and CO₂ emissions: The Bangladesh experience. IOSR Journal of Economics and Finance (IOSR-JEF), 5(6), 36–41.

Ameen

R. F. M.

, & Mourshed

(2017). Urban environmental challenges in developing countries—A stakeholder perspective. Habitat International, 64. https://doi.org/10.1016/j.habitatint.2017.04.002

Aminu Aliyu

(2005, June 9–11). Foreign direct investment and the environment: Pollution haven hypothesis revisited. Paper prepared for the Eight Annual Conference on Global Economic Analysis, Lübeck, Germany.

Balaguer

, & Cantavella

(2016). Estimating the environmental Kuznets curve for Spain by considering fuel oil prices (1874–2011). Ecological Indicators, 60, 853–859.

Beak

, & Koo

W. W.

(2009). A dynamic approach to FDI-Environment nexus: The case of China and India. Journal of International Economic Studies, 13(2), 1598–2769.

Belloumi

(2014). The relationship between trade, FDI and economic growth in Tunisia: An application of the autoregressive distributed lag model. Economic Systems, 38(2), 269–287

10.

Blackburne

E. F.

, & Frank

M. W.

(2007). Estimation of nonstationary heterogeneous panels. The Stata Journal, 7(2), 197–208.

11.

Boluk

, & Mert

(2015). The renewable energy, growth and Environmental Kuznets Curve in Turkey: An ARDL approach. Renewable and Sustainable Energy Reviews, 52(C), 587–595.

12.

Bowonder

(1985). The Bhopal accident. Environmentalist, 5, 89–103.

13.

Cai

, Che

, Zhu

, Zhao

, & Xie

(2018). Will developing countries become pollution havens for developed countries? An empirical investigation in the Belt and Road. Journal of Cleaner Production, 198(5), 624–632. https://doi.org/10.1016/j.jclepro.2018.06.291

14.

Choi

, Heshmati

, & Cho

(2010). An empirical study of the relationships between CO₂ emissions, economic growth and openness (IZA Discussion Paper No. 5304). http://ftp.iza.org/dp5304.pdf

15.

Dasgupta

, Laplante

, Wang

& Wheeler

(2002). Confronting the environmental Kuznets curve. Journal of Economic Perspectives, 16(1), 147–168.

16.

Grossman

G. M.

, & Krueger

A. B.

(1991). Environmental impacts of a North American Free Trade Agreement (NBER Working Papers Series No. 3914). NBER.

17.

Grossman

G. M.

, & Krueger

A. B.

(1995). Economic growth and the environment. Quarterly Journal of Economics, 110, 353–377.

18.

Gunarto

(2020). Effect of economic growth and Foreign direct investment on carbon emission in the Asian States. International Journal of Energy Economics and Policy, 10, 563–569. https://www.econjournals.com/index.php/ijeep/article/view/10218

19.

Gupta

, & Yadav

(2014). Understanding CO₂ emissions in India: EKC and beyond. Presented at Fifth World Congress of Environmental and Resource Economists.

20.

Hailemariam

, Dzhumashev

, & Shahbaz

(2019). Carbon emissions, income inequality and economic development. Empirical Economics, 59, 1139–1159.

21.

Hill

R. J.

, & Magnani

(2002). An exploration of the conceptual and empirical basis of the environmental Kuznets curve. Australian Economic Papers, 41(2), 239–254.

22.

Hitam

, & Borhan

(2012). FDI, growth and the environment: Impact on quality of life in Malaysia. Procedia – Social and Behavioral Sciences, 50, 333–342. https://doi.org/10.1016/j.sbspro.2012.08.038

23.

K. S.

, Pesaran

M. H.

, & Shin

(2003). Testing for unit roots in heterogeneous panels. Journal of Econometrics, 115(1), 53–74.

24.

Jayanthakumaran

, Verma

, & Liu

(2012). CO₂ emissions, energy consumption, trade and income: A comparative analysis of China and India. Energy Policy, 42, 450–460.

25.

Kahouli

, Omri

, & Chaibi

(2014). Environmental regulations, trade, and foreign direct investment: Evidence from Gravity Equations (IPAG Working Paper No. 2014-189). Department of Research, IPAG Business School.

26.

Khan

M. A.

, Khan

M. Z.

, Zaman

, & Arif

(2014). Global estimates of energy-growth nexus: Application of seemingly unrelated regressions. Renewable and Sustainable Energy Review, 29, 63–71.

27.

Kheder

S. B.

(2010). French FDI and pollution emissions: An empirical investigation. University of Paris. http://citeseerx.ist.psu.edu/viewdoc/download;jsessionid=5BB5F009EC31E8C646843C41B6A6BC7F?doi=10.1.1.193.8893&rep=rep1&type=pdf

28.

Kuznets

(1955). Economic growth and income inequality. The American Economic Review, 45(1), 1–28.

29.

, Wang

, & Zhao

(2016). Environmental Kuznets curve in China: New evidence from dynamic panel analysis. Energy Policy, 91, 138–147.

30.

Liang

F. H.

(2006). Does foreign direct investment harm the host country’s environment? Evidence from China. SSRN Electronic Journal. https://dx-doi-org.web.bisu.edu.cn/10.2139/ssrn.1479864

31.

Linh

D. H.

, & Lin

S. M.

(2012). CO₂ emissions, energy consumption, economic growth and FDI in Vietnam. Managing Global Transitions, 12(3), 219–232.

32.

Maddala

G. S.

, & Wu

(1999). A comparative study of unit root tests with panel data and a new simple test. Oxford Bulletin of Economics and Statistics, 61, 631–652.

33.

Neagu

(2019). The link between economic complexity and carbon emissions in the European Union Countries: A model based on the environmental Kuznets curve (EKC) approach. Sustainability, 11(17), 1–27.

34.

C. F.

, Choong

C. K.

, & Lau

L. S.

(2020). Environmental Kuznets curve hypothesis: Asymmetry analysis and robust estimation under cross-section dependence. Environmental Science and Pollution Research, 27, 18685–18698. https://doi.org/10.1007/s11356-020-08351-w

35.

Nunnenkamp

(2001). Foreign direct investment in developing countries: What policymakers should not do and what economists don’t know (Kiel Discussion Papers 380). Kiel Institute for the World Economy (IFW).

36.

Omri

, Nguyen

D. K.

, & Rault

(2014). Causal interactions between CO₂ emissions, FDI, and economic growth: Evidence from dynamic simultaneous equation models. Economic Modelling, 42, 382–389.

37.

Panayotou

(1993). Empirical tests and policy analysis of environmental degradation at different stages of economic development (Working Paper WP238). Technology and Employment Programme, International Labour Office.

38.

Pedroni

(1999). Critical values for cointegration tests in heterogeneous panels with multiple regressors. Oxford Bulletin of Economics and Statistics, 61(S1), 653–670.

39.

Pedroni

(2000). Fully modified OLS for heterogeneous cointegrated panels. Advances in Econometrics, 15, 93–130.

40.

Pedroni

(2004). Panel cointegration: Asymptotic and finite sample properties of pooled time series tests with an application to the PPP hypothesis. Econometric Theory, 20, 597–625.

41.

Pesaran

M. H.

, & Smith

(1995). Estimating long-run relationships from dynamic heterogeneous panels. Journal of Econometrics, 68(1), 79–113.

42.

Pesaran

M. H.

, & Shin

(1999). An autoregressive distributed lag modelling approach to cointegration analysis. In Strom

(ed.), Econometrics and economic theory in the 20th century: The ragnar frisch centennial symposium. Cambridge University Press.

43.

Pesaran

M. H.

, Shin

& Smith

(1999). Pooled mean group estimation of dynamic heterogeneous panels. Journal of the American Statistical Association, 94(446), 621–634.

44.

Rahman

, Woahid Murad

S. M.

, Ahmad

, & Wang

(2020). Evaluating the EKC hypothesis for the BCIM-EC member countries under the belt and road initiative. Sustainability, 12(4), 1–20.

45.

Rothman

D. S.

(1998). Environmental Kuznets curve: Real progress or passing the buck? A case for consumption-base approaches. Ecological Economics, 25(2), 177–194.

46.

Samargandi

, & Fidrmuc

, & Ghosh

(2015). Is the relationship between financial development and economic growth monotonic? Evidence from a sample of middle-income countries. World Development, 68(C), 66–81.

47.

Seker

, Ertugrul

H. M.

, & Cetin

(2015). The impact of foreign direct investment on environmental quality: A bounds testing and causality analysis for Turkey. Renewable and Sustainable Energy Reviews, 52(C), 347–356

48.

Selden

T. M.

, & Song

D. Q.

(1994). Environmental quality and development: Is there a Kuznets curve for air pollution emission. Journal of Environmental Economics and Management, 27(2), 147–162.

49.

Shahbaz

, Balsalobre-Lorente

, & Sinha

(2019). Foreign direct investment–CO₂ emissions Nexus in Middle East and North African countries: Importance of biomass energy consumption. Journal of Cleaner Production, 217, 603–614, ISSN 0959-6526. https://doi.org/10.1016/j.jclepro.2019.01.282

50.

Shahbaz

, Nasreen

, Abbas

, & Anis

(2015). Does foreign direct investment impede environmental quality? New evidence from high, middle and low income countries. Energy Economics, 51, 275–287.

51.

Shahbaz

, Nasreen

, & Afza

(2011). Environmental consequences of economic growth and foreign direct investment: Evidence from panel data analysis. Bulletin of Energy Economics, 2(2), 14–27

52.

Sharma

, Kautish

, & Uddin

G. S.

(2020). Do the international economic endeavors affect CO₂ emissions in open economies of South Asia? An empirical examination under nonlinearity. Management of Environmental Quality, 31(1), 89–110. https://doi.org/10.1108/MEQ-05-2019-0099

53.

Stavropoulos

, Wall

& Xu

(2018). Environmental regulations and industrial competitiveness: Evidence from China. Applied Economics, 50, 1378–1394. https://doi.org/10.1080/00036846.2017.1363858.

54.

Tamazian

, & Bhaskara

R. B.

(2010). Do economic, financial and institutional developments matter for environmental degradation? Evidence from transitional economies. Energy Economics, 32(1), 137–145.

55.

Tiwari

A. K.

, Shahbaz

, & Hye

Q. M.

(2013). The environmental Kuznets curve and the role of coal consumption in India: Cointegration and causality analysis in an open economy. Renewable and Sustainable Energy Reviews, 18(C), 519–527.

56.

A. H.

, Ha

D. T.

, Nguyen

H. M.

, & Vo

D. H.

(2019). The impact of foreign direct investment on environment degradation: Evidence from emerging markets in Asia. International Journal of Environmental Research and Public Health, 16(9), 1636. https://doi.org/10.3390/ijerph16091636

57.

Tutulmaz

(2015). Environmental Kuznets curve time series application for Turkey: Why controversial results exist for similar models? Renewable and Sustainable Energy Reviews, 50(C), 73–81.

58.

U. T. S. C., & Mitra

(2020). Development and degradation: The Nexus between GDP, FDI, and pollution in India. Emerging Economy Studies, 6(1), 39–49. https://doi.org/10.1177/2394901520907712

59.

Ullah

, & Awan

(2019). Environmental Kuznets curve and income inequality: Pooled mean group estimation for Asian developing countries. Forman Journal of Economic Studies, 15, 157–179. 10.32368/FJES.20191507

60.

UNCTAD. (2014, January). Global investment trends monitor.

61.

Wang

S. X.

, Fu

Y. B.

, & Zhang

Z. G.

(2015). Population growth and the environmental Kuznets curve. China Economic Review, 36(C), 146–165.

62.

Zakarya

G. Y.

, Mostefa

, & Abbes

S. M.

(2015). Factors affecting CO₂ emissions in the BRIC countries: A panel data analysis. Procedia Economics and Finance, 26, 114–125.