第 55 卷 第 1 期
2020 年 2 月
JOURNAL OF SOUTHWEST JIAOTONG UNIVERSITY
Vol. 55 No. 1
Feb. 2020
ISSN: 0258-2724 DOI：10.35741/issn.0258-2724.55.1.49
Research Article Economics
E
NERGY
C
ONSUMPTION
,
I
NCOME
,
T
RADING
O
PENNESS
,
AND
E
NVIRONMENTAL
P
OLLUTION
:
T
ESTING
E
NVIRONMENTAL
K
UZNETS
C
URVE
H
YPOTHESIS
能源消耗，收入，交易开放度和环境污染：测试环境库兹涅茨曲
线假说
Nguyen Van Chien Thu Dau Mot University
6, Tran Van On Street, Phu Hoa Ward, Thu Dau Mot City, Binh Duong, Vietnam, [email protected],
Abstract
The purpose of this study is to examine the effects of energy consumption, economic growth, and trade openness on environmental pollution in a developing country, especially in the case of Vietnam. The study was conducted on the basis of time-series data collected in the period of time between the years 1990 and 2014. By a method of autoregressive distributed lag and testing the hypothesis of the environmental Kuznets curve, our result demonstrated that environmental Kuznets curve could be found in both the long run and short run. There existed an inverted U-shaped relationship between different pollutants and per capita income. Further, energy consumption could positively affect carbon dioxide (CO2) emissions in the short run, but negatively could affect CO2 emissions in the long run because of transformation from non-renewable energy sources to renewable energy sources. In addition, environmental pollution converged on its long-run equilibrium by at least 29.4% with the speed adjustment via the channel of income, energy consumption, and trade openness. In terms of trade openness, the country has a positive and significant effect on CO2 emissions in both the long run and short run.
Keywords:Environmental Pollution, Energy, Environmental Kuznets Curve, Inverted U-Shaped Curve, Trade Openness 摘要 这项研究的目的是检验发展中国家的能源消耗，经济增长和贸易开放程度对环境污染的影响 ，尤其是在越南。该研究是基于 1990 年至 2014 年之间的时间序列数据进行的。采用自回归分布 滞后方法并检验了环境库兹涅茨曲线的假设，我们的结果表明环境库兹涅茨曲线可以从长远来看 都可以找到。不同污染物与人均收入之间呈倒 U 型关系。此外，由于从不可再生能源向可再生能 源的转变，能源消耗在短期内可能对二氧化碳排放产生积极影响，但从长远来看可能对二氧化碳
排放产生负面影响。此外，通过收入，能源消耗和贸易开放渠道的速度调整，环境污染在其长期 均衡上至少收敛了 29.4％。在贸易开放方面，无论从长期还是短期来看，该国都对二氧化碳排放 产生积极而重大的影响。
关键词: 环境污染，能源，环境库兹涅茨曲线，倒 U 形曲线，贸易开放
I. I
NTRODUCTIONIn the industrial revolution and nowadays, environmental pollution (EP) has become such an important problem and increasingly affected both the developed and developing countries in general and damaged health and human being in specific.
Pollution releases from any form of substance (such as solid, liquid, or gas), as well as any form of energy (such as heat, sound, noise, or
radioactivity) into the lives. In the 21st century,
the fourth industrial revolution has been
conducted and transformed to use more renewable energy and ensure an eco-friendly
environment. Modern societies are also
concerned about environmental protection, in particular, to reduce negative externalities on the environment, wildlife, and human lives.
Since launching economic reform in
1986, Vietnam has rapidly transformed both the political and economic situations. Such a rapid economic based on a socialist-oriented market economy, Vietnam has supported numerous free trade agreements in order to receive foreign investment and trade [1].
It is evident that the environmental quality of a country is predominantly influenced by a
combination of energy consumption and
economic growth (EG). In the past, most non-renewable energy sources had been consumed because of low price, and more convenience. As a result, in India, the effect of EG on pollution is dependent on the kind of energy use [2]. In addition to habit and income in households, more per capita incomes are likely to pay attention to the consumption of more renewable energy sources, because environmental pollutants have various adverse effects on health meanwhile these renewable energy sources can protect the environment and ensure the families’ health [3], [17], [18].
Consequently, because of the increasing pollution in recent years, many empirical studies
have predominantly compared numerical
modeling of the factors affecting EP [4], [5]. In this regard, Vietnam was considered to have the potential to achieve more economic development, but carbon dioxide (CO2) emissions in the country have significantly damaged the lives of
human beings and society. More specifically, environmental quality in Vietnam has been worsened over recent years. Based on the World Development Indicators (WDI), such a large value of CO2 emissions was released into the environment from 0.263 metric tons per capita in 1989 to 1.819 metric tons per capita in 2015.
For reasons discussed above, this study aims to examine the determinants of EP in both the short run and long run, as well as to test the hypothesis of the environmental Kuznets curve (EKC), particularly in Vietnam. Factors are energy consumption (EC), EG, trade openness (TO), and CO2 emission. The objectives of the present work are as follows: (i) to examine EC, EG, and TO and its impact on EP and (ii) to discuss the major conclusion in the case of Vietnam.
Theoretically, energy use, EG achievement, and TO are the leading factors that could affect environmental degradation. This empirical work is a significant contribution to the literature review of the factors affecting environmental degradation in a specific case in Asia. In addition, this study may significantly provide information to all, especially to the policy makers, researchers, and government in order to protect the environment.
The rest of the paper is organized as follows: Section 2 presents the literature review, whereas Section 3 discusses the data and data sources, as well as methodology development and techniques used in the study. Further, Sections 4 and 5 depict
the results. Finally, Section 6 states the
conclusion and policy recommendations.
II. T
HEORETICALR
EVIEW ANDF
RAMEWORKIn the background of economic development regarding environmental protection, the five key global environmental indicators are as follows: biological diversity; food production; average
global surface temperature and CO2
concentrations in the atmosphere; human population; and resource depletion. Numerous empirical studies have examined the existence of the inverted U-shaped curve regarding the relationship between income level and EP
indicators. Polluted indicators may understand as
CO2 emissions, carbon monoxide (CO)
emissions, wastewater discharge, sulfur dioxide (SO2) emissions, suspended particles, and smog in the air. Theoretically, Kuznets had studied and developed a hypothesis that the link between environmental quality and the level of economic development can be significantly appeared. The numerous environmental indicators tend to worsen the issue, when the level of economic development happens until the threshold of income, and fall with increasing income per capita. EKC was checked in a few developing and developed countries across the world [6], [7]. A recent study showed that some emerging countries are facing lots of environmental issues in the process of economic development. EG, urbanization, and in particular industrialization has been considered as the major reasons for EP in many countries, such as China [8], Germany, the USA [5], and Turkey [7]. There have been a number of studies that examined the determinants of EP, such as the studies of Yang et al. [8], Cai et al. [5], and Pata [7]. The linkages among factors are focused on the long run and short run [4], [5], [9], [10].
In Turkey, Pata [7] used an autoregressive
distributed lag (ARDL) approach and
cointegration tests in the period between 1974 and 2014. Results showed that factors such as EG, urbanization, and financial development could increase EP, while the use of renewable energy, hydropower, and clean energy did not affect CO2 emissions. As a result, the use of more renewable energy did not affect the reduction of CO2 emissions. Further, the study also supported the hypothesis of EKC in Turkey, as well as an inverted U-shaped link between EG achievement and environmental quality. This hypothesis was further supported in a specific sector; an inverted U-shaped relationship between growth in the industrial sector and EP can be found based on the studies of Muhammad [6] and Abokyi et al. [11].
The study of Muhammad [6] was conducted on data of 68 countries, i.e., developed, developing, and emerging countries, as well as the Middle East and North Africa (MENA) countries, which used panel data and techniques of generalized method of moments (GMM), and system GMM during 2001-2017 period. The evidence in developed and MENA countries supported the EKC hypothesis; however, the level of environmental quality could worsen in emerging economies because of the increase in EG. In addition to ECs, emissions of CO2 will be certainly increased in countries with EC growth.
Thus, environmental quality can be improved in the context of using more environmentally friendly technologies.
As suggested by Mikayilov et al. [12], using the secondary data in the years between 1992 and 2013 in Azerbaijan, especially to analyze the relationship between EG and CO2 emissions in order to test the EKC hypothesis, and EKC is not found in Azerbaijan. Economic development and industrialization in the country had led to increase in energy consumption and CO2 emissions. To reduce environmental degradation and negative externalities from polluted effects, the country needs to be more effective in energy use and support the instruments of carbon pricing in the manufacture, trade, and social awareness.
By focusing on a few developed economies, Cai et al. [5] conducted a study on Germany and
the USA. The study indicated that a
unidirectional impact was run from the use of clean energy on emissions of CO2 in both countries. Moreover, this study suggested that the two countries should use energy efficiency in order to reduce environmental degradation. Wasti and Zaidi [4] examined this issue in Turkey by using the ARDL model; they also reviewed stationarity based on the tests of Augmented Dicky-Fuller (ADF) and Phillips-Perron (PP). Significantly, there exists a two-way relationship between environmental quality and energy use. Moreover, they further demonstrated that a unidirectional effect was pursued from the EG achievement to CO2.
Regarding trade, various empirical studies have been focused on the relationship between TO and CO2 emissions. As suggested by
Mutascu [13], by using the method of
wavelet tool over the 1960–2013 period in France, it is obvious to confirm that TO positively generates CO2 emissions, but this effect happens at low frequency. Similarly, Sannassee and Seetanah [9] also supported this evidence in the case of a developing country of East Africa (Mauritius) in the years 1976–2013. By more contribution, trade has positively and significantly affected the increase of CO2 in both the long run and short run. Sannassee and Seetanah [9] deeply discussed that the growth in the manufacturing sector had been the main contributor to environmental degradation. Sebri and Ben-Salha [10] showed that there was the existence of long-run equilibrium relationships among EG, TO, renewable energy use, and CO2 emissions in Brazil, Russia, India, and China (BRICS) countries.
III. R
ESEARCHM
ETHODOLOGY A. Data SourcesBecause of the availability of the secondary data until 2014, this study was estimated from 1985 to 2014. All of the data were collected from the General Statistics Office of the Ministry of Planning and Investment in Vietnam, Ministry of Finance, WDI, and United Nations Conference on Trade and Development (UNCTAD) statistics. The data includes EP, EG, EC, and TO.
B. Methodology
The impact of EG, EC, and TO on EP has been examined in a number of developed and developing countries, as well as countries in transition. Theoretically, EKC demonstrates that an inverted U-shaped curve between different pollutants and per capita income could be found. It is related to a relationship of a higher per capita income and worse environment quality.
Followed by studies of Wasti and Zaidi [4], Cai et al. [5], Yang et al. [8], and Sannassee and Seetanah [9], as well as other previous studies, the original model equation is written as follows:
Y = function (X1, X2, X3, Xi…Xn) (1)
The study focused on the relationship of EG, EC, TO, and EP; the following model is given by:
lnEPt = α0 + α1 ln ECt + α21 ln EGt + α3 TOt +
εt (2)
Based on the EKC hypothesis, the estimation equation is specifically given as follows:
lnEPt = α0 + α1 ln ECt + α21 ln EGt + α22 ln
EG2t + α3 lnTOt +εt (3)
The ARDL approach, corresponding to
Equation (3) with lnEPt as dependent variable, is
given as follows:
(4) The bounds test is based on Wald’s statistics.
The null hypothesis H0: There is no
relationship
The alternative hypothesis Ha: There is
cointegration
If there is cointegration, the error correction model (ECM) is specifically given as follows:
(5) where:
is denoted for the error correction term and extracted residuals from the regression of the long-run equation. This indicator is denoted for the long-run parameter.
is the speed of the adjustment parameter with a negative sign. This factor is expected to be between -1 and 0, which reflects that the system is converged to equilibrium and further indicates that the estimated model is stable.
The subscript t indicates the time period (t = 1985… 2014)
α0, α1, α21, α22, and α3 are estimation
coefficients.
εt is an error in year t.
lnEPt = is the dependent variable, indicating
the level of EP, and estimated by the natural logarithm of CO2 emission per capita (metric tons).
lnECt = is the independent variable, indicating
EC, and calculated by the natural logarithm of electricity power consumption in kWh per capita.
lnEGt = is the independent variable, indicating
EG, and calculated by the natural logarithm of gross domestic product per capita.
lnEG2t = is the independent variable and
calculated by the square of the natural logarithm of per capita income. This variable is used for testing the EKC hypothesis.
lnTOt = is the independent variable, indicating
TO, and calculated by the natural logarithm of TO (% of GDP).
IV. R
ESEARCHR
ESULTS A. Descriptive StatisticsTable 1 depicts the descriptive statistics of the variables used in this study regarding their mean and standard deviation (SD), as well as the
results indicated that environmental quality in Vietnam has been worsened in recent years. In fact, a large value of CO2 emissions was released into the environment from 0.263 metric tons per capita in 1989 to 1.819 metric tons per capita in 2015. In addition to EC, the volume of EC was significantly changed at a huge increase between 260.791 kg of oil equivalent per capita in 1991 and 669.699 kg of oil equivalent per capita in 2010. Further, EG on average in Vietnam was maintained a high rate of 7% during the period of 1990-2018. The per capita income experienced the highest level of 2030.262 US dollars in 2014 and the lowest of 558.5196 in 1989. TO in the economy had steadily expanded from 28.95% in 1988 to 169.53% in 2015. As a result, TO indicated a relative scale of foreign trade in Vietnam’s economy and measured in total import-export value against GDP. In recent years, Vietnam’s economy has been a very high degree of TO because of expansion in trade policy, which means that the country may be more vulnerable to external economic fluctuations.
Table 1.
Descriptive statistics
Var Obs Mean SD Min Max
EP 30 0.838 0.518 0.263 1.819 EC 30 411.941 150.732 260.791 669.699 EG 30 654.784 558.519 94.564 2030.262 TO 30 102.844 46.154 18.950 169.534
B. Unit Root Test
Based on the ADF test, we have:
The null hypothesis H0: The time series
is non-stationary.
The alternative hypothesis Ha: The time
series is stationary.
Performing the ADF test, Table 2 shows that the result obtained from the ADF test indicates that all series are stationary, except lnEP. For the
first difference of lnEP, it is stationary at a 5% significance level.
Table 2.
Unit root test results
Var Order of integration ADF test Hypothesis lnEP I(0) I(1) 0.1662 0.0140
Null hypo. is not rejected Null hypothesis is rejected lnEC I(0) 0.0939 Null hypothesis is rejected lnEG I(0) 0.0002 Null hypothesis is rejected lnEG2 I(0) 0.0000 Null hypothesis is rejected lnTO I(0) 0.000 Null hypothesis is rejected Note: This table shows the results of the stationary test, which includes the ADF test.
C. Appropriate Lag Length for ARDL
In practice, based on the optimal lag length, the selection of an appropriate model is significantly necessary. There are several criteria for choosing the optimal lag length in a time series, i.e., likelihood ratio (LR), Akaike information criterion (AIC), Schwarz Bayesian criterion (SBC), final prediction error (FPE), and Hannan-Quinn criterion (HQ).
SBC indicated that the optimal lag length is one. Using LR, FPE, AIC, and HQ indicated that the optimal lag length is two and also more popular in all criteria. Furthermore, Wooldridge [14] indicated that with annual data, the number of lags is typically small, one or two lags in order not to lose degrees of freedom. In particular, in times-series data with limited observations, it is often preferred to use AIC in selecting the lag length. In conclusion, Table 3 shows that the number of optimal lags is two lags.
Table 3.
Appropriate lag length for ARDL
Lag LR FPE AIC SBC HQ
0 NA 3.5e-10 -10.39 -10.20 -10.33 1 193.06 2.0e-13 -17.88 -16.91* -17.60 2 32.66* 1.1e-13* -18.57* -16.83 -18.07* 3 13.40 1.8e-13 -18.37 -15.85 -17.64
D. ARDL Bounds Test of Cointegration
To perform the bounds test for cointegration, the conditional ARDL (p, q1, q2) model with four variables was specified. Based on the optimal lag length (2 lags), the study is to analyze the bounds test. Theoretically, ARDL bounds test can be used for cases provided none of the series
is beyond I(1). Further, the bounds test should be performed on the level form of the variables and not on the first difference
Based on the bounds test proposed by Pesaran, Shin, and Smith [15], we have:
The alternative hypothesis Ha: having cointegration.
If the calculated F-statistics are greater than the critical value for the upper bound I(1), we can conclude that there is cointegration. That is, there is a long-run relationship. Reject the null hypothesis. Further, the estimation of a long-run model, which is ECM.
If the calculated F-statistics are lower than the critical value for the lower bound I(0), we can conclude that there is no cointegration; hence, there is no long-run relationship. Do not reject the null hypothesis. Further, the estimation of a short-run model, which is the ARDL model.
Generally, lnEP is the dependent variable. It is evident that if F-statistic is lower than the critical value for the lower bound I(0) at the significance level of 1%, we can conclude that there is no cointegration, as well as no long-run relationship between EP and its impact. As a result, the estimation of a short-run model is the ARDL model.
Table 4 depicts the computed F-statistic for the ARDL Bounds Test, denoting that testing the existence of the long-run relationship between
the variables can be found or not. As a result, lnEP is the dependent variable; the calculated F-statistic (F lnEP/ lnEC, lnEG, lnEG2, lnTO) = 10.84 and is greater than upper critical bound at a 1% level of significance (the upper critical value at a 1% level of significance is 5.06). Moreover, in the case of exclusion of EKC, lnEG2 is not presented in the model; F-statistic is 9.00 and also greater than the critical value for the lower bound I(1) at the significance level of 5%. A cointegration between lnEP and its determinants can be found. In conclusion, it suggests that there is a cointegration between the lnCO2 emission and its determinants in the case of Vietnam.
Table 4.
Results of the ARDL cointegration test N.
of var
F-sta. ARDL bounds test
Dependent variable: lnEP Lower bounds I(0), Upper bounds I(1)
95% 97.5% 99%
k I(0) I(1) I(0) I(1) I(0) I(1) 4* 10.84 2.86 4.01 3.25 4.49 3.74 5.06 3** 9.00 2.86 4.01 3.25 4.49 3.74 5.06 Note: (**) and (*) denote in the case of inclusion and exclusion of the EKC characteristic, respectively.
Table 5.
Error correction representation of the selected ARDL Dependent variable: lnCO2
Regressor Short-run ADRL (1, 0, 0, 0, 2) Short-run ADRL (1, 2, 0, 2)
Coefficient P-value Coefficient P-value
ECMt-1 -0.5448* 0.000 -0.2936** 0.011 lnEPt-1 -0.5448* 0.000 -0.2936* 0.000 lnECt 0.6113*** 0.054 -0.2363 0.308 lnECt-1 1.0842* 0.003 lnECt-2 0.9089** 0.014 lnEGt 1.2041* 0.000 0.2362* 0.000 lnEG2t -0.0871* 0.000 lnTOt 0.1312* 0.006 0.1831* 0.000 lnTOt-1 0.0322 0.616 0.0169 0.773 lnTOt-2 0.2148* 0.001 0.2236* 0.000 Constant -8.4750* 0.001 -1.0192 0.403 Note: (*), (**), and (***) denote significance at 1%, 5%, and 10%, respectively.
Table 6.
Long-run coefficients of the ARDL model Dependent variable: lnCO2
Regressor Long-run ARDL (1, 2, 0, 0, 2) Long-run ARDL
(1, 2, 0, 2)
Coefficient P-value Coefficient P-value
lnEC -0.1789* 0.006 -0.8049 0.432 lnEG 2.2100* 0.000 0.8046** 0.019 lnEG2 -0.1600* 0.000
lnTO 0.2408** 0.035 0.6238** 0.026 Note: (*), (**), and (***) denote significance at 1%, 5%, and 10%, respectively.
第 55 卷 第 1 期
2020 年 2 月
JOURNAL OF SOUTHWEST JIAOTONG UNIVERSITY
Vol. 55 No. 1
Feb. 2020
E. Diagnostic Tests
The results of the diagnostic tests are shown in Table 7. First, we tested the serial correlation of the selected ARDL based on the Durbin-Watson D statistics and Breusch-Godfrey LM test; it depicts that no autocorrelation can appear in the model. Second, the heteroscedasticity test, based on the Breusch-Pagan/Cook-Weisberg test, White's test, and Cameron and Trivedi's decomposition of IM-test, indicates that the model does not exist heteroscedasticity.
Table 7. Diagnostic tests Tests A: Serial correlation Durbin-Watson D statistics DW = 1.713517, P-value = 0.7349
Breusch Godfrey LM test P-value = 0.6820 B: Heteroscedasticity Breusch-Pagan/Cook-Weisberg test chi2(1) = 0.17 Prob > chi2 = 0.6787 White's test for Ho: homoskedasticity Ha: unrestricted heteroskedasticity
chi2(27) = 28.00 Prob > chi2 = 0.4110 Cameron & Trivedi's
decomposition of IM-test
chi2 = 28 Prob > chi2 = 0.4110
The study conducted to test the stability of the ARDL model based on error correction model (ECM). By using the cumulative sum of recursive residuals (CUSUM) and particularly cumulative sum of squares of recursive residuals (CUSUMSQ), Brown et al. [16] recognized that if both plots remain within critical bounds at a 5% level of significance (Figure 1), the model is certainly stable.
Figure 1. The plot of the cumulative sum of squares of recursive residuals
V. D
ISCUSSIONRegarding the estimation results by the ARDL approach, our results show that there exist the short-run and long-run relationships between EP and other independent variables as EC, EG, EG2, and TO in the case of Vietnam, as follows:
Table 5 describes the results of the error correction representation of the selected ARDL model. The coefficients of the variables with ∆ sign indicated the short-run relationship between the level of dependent variables (EP) and independent variables (EC, EG, EG2, and TO). Results showed that in the short run, the variables, such as the coefficients of lnEG, lnEG2, lnTO,
and lnTOt-2, are significant at the 1% level.
More specifically, EKC is a hypothesized relationship between environmental quality and economic development, which denoted by EG and can be found in the case of Vietnam. Similarly, there exists an inverted U-shaped relationship between different pollutants and per capita income. Vietnam has rapidly increased in
GDP and socio-economic development.
When economic development growth occurs, the evidence confirms that the environment will certainly worsen at a certain point. This study supports the findings of Muhammad [6] in MENA countries, Pata [7] in Turkey.
Regarding EC, the evidence is found in both scenarios at a 10% level of significance, EC can positively generate CO2 emissions in the short run but negatively generate CO2 emissions in the long run. In the long run, by using EC, Vietnam has transformed from non-renewable energy sources ( i.e., fossil fuels: coal, petroleum, and natural gas) to renewable energy sources, and it will be good for the environment. It is in line with Wasti and Zaidi [4] in Turkey, as well as Cai et al. [5] and Muhammad [6] in the USA and Germany.
Furthermore, the coefficient of the error correction term is either –0.544 or -0.294, and it is significant at 1%. Theoretically, it is expected to be between -1 and 0; it implies that the system is converged to equilibrium and further indicates that the estimated model is stable. In the analysis, the level of the speed of adjustment from the previous year’s disequilibrium in EP to the current year’s equilibrium was somewhat huge and roughly 54.4% (or 29.4%). It further
indicates that EP converges on its long-run equilibrium by at least 29.4% with the speed adjustment via the channel of EG, EC, and TO.
Table 6 depicts the long-run relationship between the variables. The results indicated that the EKC effects could also be found in the long run at the 1% level of significance. In terms of TO, the study demonstrates that TO has a positive and significant impact on CO2 emissions in both the long run and short run, similar to the studies of Sannassee and Seetanah [9] in Mauritius and Mutascu [13] in France.
VI. C
ONCLUSION ANDR
ECOMMENDATIONSThe paper aimed to estimate the determinants of EP in Vietnam for the period from 1985 to 2014. Basically, the study selected this period for the research because of the availability of data used in the study.
The objective of the paper was to investigate EC, EG, and TO in the economy, which impacted on EP (CO2 emissions) in a developing country in Pacific Asia, such as Vietnam. All indicators were examined to check the unit root and cointegrated effects in the long run and short run based on the ARDL model. In this study, we tested EKC in relation to economic development in Vietnam.
Theoretically, the numerous previous studies, conducted on EKC, supported that economic development in a country can significantly lead to a deterioration in its environment. However, after a certain level of EG in the future, the environmental quality will be improved, and levels of environmental degradation will be reduced. By testing in Vietnam, we found that EKC can be found in both the long run and short run. There exists an inverted U-shaped relationship between different pollutants and per capita income. In addition, EC can positively generate CO2 emissions in the short run but negatively generate CO2 emissions in the long run because of the transformation use of energy resources from non-renewable energy sources to renewable energy sources. Further, EP converges on its long-run equilibrium by at least 29.4% with the speed adjustment via the channel of EG, EC, and TO. For TO, it has a positive and significant impact on CO2 emissions in both the long run and short run.
The findings have led to some major recommendations. Firstly, Vietnam’s government continues to transform the use of energy resources to clean EC through the promotion of renewable energy use. Secondly, inward foreign direct investment (FDI) attraction to Vietnam
needs to use technology-intensive projects instead of labor-intensive projects that can generate more pollution.
A
CKNOWLEDGMENTThe authors received no financial support for the research, authorship, and/or publication of this article.
