Đường cong kuznets về môi trường đối với khí thải carbon dioxide trường hợp các nước ASEAN

ENVIRONMENTAL KUZNETS CURVE FOR CARBON DIOXIDE EMISSIONS IN ASEAN COUNTRIES Pham Xuan Hoan, Nguyen Cam Nhung and Vu Thi Hien Thu' Abstract The environmental Kuznets curve hypothesis is a theory by which the relationship between per capita GDP and per capita pollutant emissions has an inverted U shape While impressive growth of ASEAN countries in the last few decades has significantly helped improve the people's standard of living and reduce the poverty rates, it has also exerted tremendous pressure on the environment of these countries The question is whether economic growth and higher income levels can eventually lead to a reduction in environmental pollution, as proposed by Environmental Kuznets curve (EKC) hypothesis The paper is to investigate the validity of EKC hypothesis for the case of carbon dioxide emissions in ASEAN countries (excluding Brunei), employing a panel data approach for ASEAN countries, using fixed effect and random effect modeling It is found that EKC does exist for the case of these countries' carbon dioxide emissions, but with a relatively high turningpoint income (approximately $14000) This suggests that more active policy measures should be taken in the development plans of ASEAN countries so that a decline in carbon dioxide emissions would occur sooner 'Pham Xuan Hoan, PhD., Director of Planning-Finance Department, Vietnam National University; Nguyen Cam Nhung, PhD., Lecture and Deputy Head of Intemational Finance Department, Faculty of Intemational and Business Economics, University of Economics and Business, Vietnam National University; Vu Thi Hien Thu, BA., graduated from Faculty of Intemational and Business Economics, University of Economics and Business, Vietnam National University 67 Socio-Economic Issues Keywords." Environmental Kuznets curve, carbon dioxide emission, ASEAN, economic growth, panel data Introduction The Environmental Kuznets Curve hypothesis (EKC) proposes that as the economy of a country develops the degradation of the environment increases, but when the economy reaches a specific level of income per capita, known as the turning point, pollution starts to decline This hypothesis implies that although in the early stages of development, pollution is imavoidable, in the end economic growth will be one of the solutions to the pollution problem In recent years, green-house gas issues have drawn more and more attention in developing ASEAN countiies The attempts to attract foreign investment and promote intemational trade have led to rapid growth in these countries, while at the same time resulting in increased energy consumption and carbon dioxide (CO2) emissions Therefore, it is cracial for pohcy makers to have an informed understanding of the income-C02 relationship m ASEAN countries in the context of recent economic growth Since little research has been conducted in this field for ASEAN countries, this paper is aimed to investigate empirically the validity of the Environmental Kuznets Curve hypothesis for carbon dioxide emissions for 10 ASEAN countries in the period between 1985 and 2010, using a non-linear model by applying pooled, fixed effect and random effect regressions It is interesting to know if the EKC does exist in this case, what are possible reasons and what is the -tuming point" income level? This would improve our understanding of flie greenhouse gas emissions to income relationship and allow die observer to precisely know where his/her country is located along the curve, thus enabling necessary policy adjustment The rest of die paper is organized as follows: Section reviews theory and past EKC works, section details econometric methodology; Section intetprets empirical findings and conclusion and policy implications are drawn in Section 68 Environmental Kuznets curve for carbon dioxide emissions in ASEAN countries Theory and Literature Review EKC hypothesizes the relationship between various indicators of enviroimiental degradation and income per capita The EKC hypothesis proposes that in the early stages of economic growth degradation and pollution increase, but beyond a certain level of income per capita (which will vary for different indicators) the trend reverses, so that at a tuming point of income levels economic growth actually leads to environmental improvement In other words, the environmental impact indicator is an inverted U-shaped fiinction of income per capita A number of studies have argued that the inverted U-shape relationship is not very likely to be found in carbon dioxide emissions The first reason is seen from a theoretical point of view: Unlike other air pollutants (such as nitrogen oxide or sulfur dioxide) which have local effects, carbon dioxide emissions cause problems on a global scale, and the social costs of global warming accme across both time and nations Therefore, free-rider behavior might lead to a close relationship between carbon emissions and income at all levels of per capita income, and the inverted-U relationship is less likely for carbon dioxide emissions than for the above 'tradifional' air pollutants (Arrow et al., 1995; Friedl, 2003) Moreover, in contrast to air or water pollufion, which can have immediate, identifiable local health effects, carbon dioxide emissions are locally innocuous and only impact the global environment over the long term This causes the lack of historical regulations or taxes on carbon dioxide emissions Indeed, until very recently carbon dioxide was not considered a harmful by-product of clean and efficient combustion and efforts to control carbon dioxide emissions have not come into effect until the last few years and have been limited to a few developed countries The second reason comes from analyzing consumer behavior Wealthy consumers in developed coimtries develop a demand for a cleaner environment which leads to more stringent enviroimiental regulations and to the relocation of dirty industries abroad The result of this is a decline in the levels of pollutants arising from production However, pollutants arising from consumption, such as solid waste and 69 Socio-Economic Issues carbon dioxide emissions (arisingfi-omautomobile use and the burning of fossil fiiels for the generation of electricity) continue to rise because they are easily externalized and thus not subject to regulations Carbon dioxide emissions have an impact on flie global environmem, rather tiian the immediate environment Thus, it is not tiie case that wealthy consumers become mote conscious of the global environment; ratiier, they develop a taste for a cleaner immediate environment, hence tiie regulations tiiat are enforced on finns producing pollutants with local impacts such as sulfur dioxide and particulate matter This suggests only the levels of these pollutants which cannot be easily extemalized decline with per capita income, whilst the levels of easily externalized pollutants such as solid waste and carbon dioxide continue to rise (Nahman et al., 2005) Therefore, only the development-induced changes in production seem a plausible driver for a decline in carbon dioxide emissions Hence, the existence of an inverted U-ciu-ve for carbon dioxide emissions intensity is expected to suggest that pollution reduction might occur as a natural by-product of economic development improving efficiency, especially in the use of energy (Roberts et al., 1997) This special feature of carbon dioxide emissions makes it an interesting case in EKC studies, which can provide implications for not only developed countries, but also for less developed countries Many contributions have empirically tested the existence of an EKC for C02 The first empirical study about CO2 emissions is the work for the Worid Bank by Shafic and Bondyopadhyay in 1992 The authors use ten indicators as environmental quality dependent variables and they set panel regressions using data ftom 149 countries for the period 1960-1990 According to their result, income has a significant effect on all indicators of environmental quality Most environmental indicators deteriorate initially with rising incomes but tend to improve as countries become richer One of the 70 Environmental Kuznets curve for carbon dioxide emissions in ASEAN countries exceptions to this pattern is carbon dioxide emissions which seem to rise monotonically as income rises For the case of Canada, and with respect to CO2 emissions, Lantz and Feng (2006) find that while population size has a positive effect on carbon dioxide emissions, technological or structural changes have switched its influence from negative to positive, favoring CO2 emissions intensity production for the sample period 1970-2000 For the case of Sweden, Kander (2005) estimates a decline of CO2 emissions after 1970 due to a stabilization of energy consumption which occurred as a result of the slow growth of the economy and a substantial decline in energy intensity within the industrial sector Some studies attempt to include forecasts about future trends of CO2 emissions Galeotti and Lanza (1999) use a panel data set of 110 countries for the period 19701996 Although the authors find evidence of an EKC-pattem in the case of C02emissions, they forecast that future global emissions will increase, due to the most rapid growth of per capita income in developing countries According to the authors, income and population play a key-role in carbon dioxide emissions Cole (2004) tests the case of CO2 emissions among other air and water pollutants and finds an income tuming point for CO2 outside the sample income range Lee et al (2009) find an evident inverted U- shape relationship between income and CO2 emissions in the case of the middle income American and European countries, but not in the case of the high or low income African, Asian and Oceania counfries for the time period 1960-2000 Aslanidis and Iranzo (2009) examine the case of non-OECD countries for the period 1971-1997 and they find no evidence of an EKC, while two regimes appear: a low income regime, where emissions increase with economic growth, and middle to high-income regime, which is associated with a decrease in enviroimiental degradation Methodology Socio-Economic Issues 3.1 Econometric Methodology The model used to examine the net effect of income on carbon dioxide emissions for panel data analysis in this paper is an approximated parabolic function, with per capita income and squared terms of per capita income as explanatory variables: ln(CO,)„ = a, + Y, +/?ilnYi, + /S^ (lnY)^, + £,-, (2) where ln(C(?2)» represents the logarithm of per capita carbon dioxide emission, InY,, the logarithm of per capita income while 0j^&-02 are the slopes of the model, u^ and y^ are the intercept parameters, i represents the cross-section of coimtries and t denotes the years of time series, E^^ is the error occurred by cross-section (countiies) or by years ontimeseries Under the EKC hypothesis, the signs of the coefficients in equation (2) are expected as follows: ai> 0, a2< 0; D a2C « D a) D The tuming point is y;,* = exp (— : ^ \ The paper applies the panel data analysis to detect the EKC curve between carbon dioxide emissions and income levels in ASEAN countries The tuming point of the EKC curve is calculated based on the estimation result Panel data helps to detect the dynamics of changes in short time series It provides more powerfiil regression by considering the place (spatial) and the time (temporal) dimensions of the data (Basel et al.,2011) The primary reason for using panel data is that it offers the opporhmity for controlling unobserved individual and/ or time specific heterogeneity, which may be con-elated witii the included explanatory variables Both time series and cross-section, when combined, enhance tile quality and quamity of data in ways fliat would be impossible using only one of these two dimensions Gujarati (2003) listed several 72 Environmental Kuznets curve for carbon dioxide emissions in ASEAN countries benefits for using panel data, such as it increases the precision of parameter estimates, allows us to sort out model temporal effects without aggregation bias, gives more informative data, less colinearity among variables, more efficiency, etc Hausman and Taylor (1981) revealed that by combining time-series and cross-sectional data, individual-specific unobservable effects (may be correlated with other explanatory variables) can be controlled The Hausman test is used to compare between the two estimated models, the fixed effects model (FEM) and random effects model (REM) Under the null hypothesis, both FEM and REM are consistent with REM being more efficient Under the altemative hypothesis, FEM is more efficient than REM Therefore, the rejection of the null hypothesis will suggest for the choice of FEM (Yaffee, 2003) 3.2 Data The panel data utilized in this paper covers the time span between 1985 and 2010, for 10 ASEAN countries (e.g Vietnam, Laos, Cambodia, Thailand, Myanmar, Indonesia, Malaysia, Singapore and Brunei) All observations are in annual frequency The reason for the observation from the year 1985 onwards is that by this year most ASEAN countries have opened to the world market, carried out economic transformation and experience dramatic growth This is compatible with the purpose of the study that is to examine the impact of economic growth on the environment The dependent variable - carbon dioxide emissions (CO2) is measured in mefric tons per capita The data is collected from "World Development Indicators" on the World Bank website The proxy of economic growth, real gross domestic product (GDP) per capita is measured by 2005 US dollars The data on this is taken from the UNCTAD statistics The missing data on this is obtained from the United Nation statistical databases Results 73 Socio-Economic Issues We shall begin the analysis by presenting some descriptive statistics of variables As can be seen ftom tables below the differences between minimum and maximum values are ratiier big for both carbon dioxide emissions and GDP per capita Table 1-1 Summary Statistics, 1985 - 2010, ASEAN countries Variable Per capita carbon dioxide emissions Mean Std Dev Min Max 4.1512 6.3319 0.047 27.274 5.2075 8.9440 0.0792 44.7037 260 Observation Emissions measured in metric tons GDP per capita measured in 2005 US dollars THE results of regression analysis of the pooled OLS model are presented in Table 1-1, and Table 1-2 represents resuhs of a regression analysis of fixed effect and random effect models For tire three models, all estimated results reveal that linear term of income per capita and non-linear term of income per capita have positive and negative effects respectively on carbon dioxide emissions at significant levels However, resuhs from pooled OLS are inappropriate for discussion, since regular OLS regression does not consider heterogeneity across countries or time, and to use this model, we need to assume homoskedasticity and no serial correlation in the data Both of these assumptions are too restrictive for the selected model Table 1-2 Pooled OLS Coefficient t-Statistic P-value -9.7643»" -7.53 0.000 InV 1.7651*" 5.08 0.000 (InY)' -0.0519** -2.29 0.023 C R-squared 0.8599 F-Statistic 487.76*** Note: *** indicates significant at 1% level, ' indicates significant at 5% level, indicates significant at 10% level 74 Environmental Kuznets curve for carbon dioxide emissions in ASEAN countries Table 1-3 Fixed an i Random Effect Models Random Effect Model Fixed Effect Model Coefficient t-Statistic P-value Coefficient t-Statistic P-value c -8.4862*** -3.69 0.005 -8.4413*** -3.40 0.001 InY 1.9826** 2.84 0.019 1.8979*** 2.59 0.009 (InY)' -0.1039* -2.00 0.076 -0.0936* -1.77 0.07 R-squared Hausman Test 0.7850 0.8201 p = 0.000 Note: *** indicates significant at 1% level, ** indicates significant at 5% level, * indicates significant at 10% level The result from the Hausman test suggests that the fixed effect model is more apt than random effect model (with p-value = 0.00 < 0.05) From the results of fixed effect model, it can be seen that the linear and square terms of logarithm real per capital income have positive and negative effect on logarithm carbon emissions at significant level of 5% and 10%, respectively, confirming the existence of an inverted U-shaped relationship between economic growth and environmental degradation Using the estimated coefficient values obtained from fixed effect model, the income tuming point for CO2 emissions in ASEAN counfries can be calculated as followed: Y(TP) = e x p ( - ^ ) = ^^p{-7:7^^^= 13917.533 (2005 US dollars) According to this calculation in comparison with data on income per capita of 10 ASEAN countries, it can be seen that only Singapore and Brunei have reached the tuming point and started to see declining carbon emissions While on the other hand, the 75 Socio-Economic Issues other eight countries' carbon emissions are expected to continue to rise until they reach a comparable level of economic development and these countiies might need a long period of growth before they can acmally reduce their carbon emissions Discussion The existence of EKC for CO2 in this case is against the idea of Arrow et al (1995) and Friedl and Getzner (2003), who propose that since carbon dioxide emissions have an impact on a global scale, not local scale, free-rider behavior might lead to the inverted-U relationship being less likely for carbon dioxide emissions than for the -traditional' air pollutants, such as sulfur dioxide and nitrogen oxide In addition, it is worth noting that the existence of EKC for carbon emissions in ASEAN countiies should not be misinterpreted as ajustification for policy inaction, nor lead to a belief that economic growth will inevitably become a solution for all environmental problems The panel data study is based on the assumption that the income effect is the same for all countries (Aslanidis et al., 2009) Therefore, there is no guarantee for all countries that economic growth will automatically lead to environmental improvement The tuming point calculated in this study ($13917.533) is relatively high for most ASEAN countries This figure suggests that of tiie 10 ASEAN countiies, only Singapore and Brunei are standing on the right hand side of the ciu-ve, which means their carbon emissions are declining with income, while the other countries with much lower income level are progressing on the left hand side of tiie curve Since tile income levels of most of these countries are still much lower than the tuming point income level (i.e.tiiesecoimtiies are standing at the lower left hand side part of tiie curve), it can be expected that their carbon dioxide emissions will continue to rise at high rates if no environmental efforts are made It can be seen tiiat Vietnam, in particular, is among die countries that are still progressing on flie left hand side of the U-curve In 2012, Vietnam's real GDP per 76 Environmental Kuznets curve for carbon dioxide emissions in ASEAN countries capital was $1567 (millions of US dollars at 2005 constant price) Therefore, it might take decades for Vietnam to reach the tuming point of $13917.533 so tiiat the country can reduce its carbon dioxide emissions However, this should be interpreted as an urge for more environmental awareness among policy makers Vietnam can be a pioneer in low-carbon development if effective technological transfer policies are implemented and low-carbon investment are encouraged Conclusions This study has shown evidence to support the EKC hypothesis for carbon dioxide emissions in the ten ASEAN countries and gives possible explanations for this Using the reduced-form of the EKC equation, it is calculated that the tuming point income level of ASEAN countries for carbon dioxide emission Is roughly $14,000 per capita This result provides a projection that on average, ASEAN counfries may be able to reduce the levels of carbon dioxide emissions - the main gas contributmg to global warming - after they reach the real income per capita of approximately $14,000 - a relatively high income level compared to most of the developing ASEAN countries This figure suggests that of the 10 ASEAN countries, only Singapore and Brunei are standing on the right hand side of the curve, which means their carbon emissions are declining with income, while the other countries with much lower income levels are progressing on the left hand side of the curve Since the income levels of most of these countries are still much lower than the tuming point income level (i.e these countries are standing at the lower left hand side part of the curve), it can be expected that their carbon dioxide emissions will accelerate at high rates if no environmental efforts are made Since it takes a long time for these countries to grow to such income levels, there are also possibilities that by the time these countries reach the tuming point income level for CO2, the environmental damage may have surpassed the recovery threshold Therefore, these countries should begin to pay attention to reducing carbon dioxide emissions and put in place a policy and regulatory framework for low-carbon growth before damage to the environment becomes irrecoverable 77 Socio-Economic Issues Overall, the paper has made a modest effort to investigate the COz-income relationship in ASEAN coimtries However, it suffers from a number of drawbacks The major shortcoming is associated with the reduced form of the EKC equation used for this empirical investigation The reduced-form approach to the investigation of the relationship between income levels and environmental quality has been a useful first step towards answering the question of how economic growth affects the environment This approach spared us the more difficult specification of structural equations and the more demanding data requirements of a more analytic approach, while it gave us a better understanding of the net effect of income on the environment However, it is unclear which factors associated with higher income level actually affect the environment Therefore, clear-cut pohcy implications for individual coimtries may not be possible using this approach Since the existence of an EKC relationship across nations today would not guarantee that global environmental degradation would decline automatically with time and economic growtii (Stem et al., 1996), the existence of an EKC relationship in ASEAN it is not Ukely to guarantee that the carbon emission levels will decline with time and economic growtii for all countries Moreover, tiie high turning point for ASEAN countries is a warning sign that bytiletimethese countries achieve this income level, tiie environmental consequences of growth may have become too severe Therefore, an active approach toward environmental policies is vital for these countries to achieve more sustainable growth Witii tiie accessibility of high technology today, developing countiies in ASEAN can leam from developed countries and gradually reducetiiecarbon dioxide emissions while at the same time achieve economic growth References Arrow, K, Bolm, B., Costanza, R., Folke, C, Holling, C.S., Janson, B., Levin, S., Maler, K, Penings, C and Pimental, D (1995), "Economic Growth Caro,ing Capacity, and the Environment Science," Vol 15, pp - 95 (reprint in Ecological Economics) 78 Environmental Kuznets curve for carbon dioxide emissions in ASEAN countries Asian Development Bank (ADB), United Nations Economic and Social Commission for Asia and the Pacific (ESCAP) and United Nations Environment Programme (UNEP) (2010), "Green Growth Resources, and Resilience" http://www.unescap.org/esd/environment/fl agpubs/GGRAP Friedl, B., Getsner, M (2003), "Determinants ofC02 Emissions in a Small Open Economy", Ecological Economics 45, pp 133-148 Nahman, A., Antrobus, G (2005), "The Environmental Kuznets curve: A Literature Review", South African Joumal of Economics 73, pp 10-35 Roberts, J.T., Grimes, P.E (1997), "Carbon Intensity and Economic Development 1962-91: A Brief Exploration of the Environmental Kuznets Curve," 25(2), pp 191198 Shafic N., Bandyopadhyay S.,(1997) "Economic Growth and Environmental Quality Time- series and cross country evidence" Policy Research working paper no 904, Worid Development Report 1992, The Worid Bank Lantz V., Feng Q., {20Q6)"Assessing income, population, and technology impacts on C02 emissions in Canada: Where's the EKC? " Ecological Economics 57, 2006, pp 229-238 Kander A., Baumol's disease and dematerialization of the economy Ecological Economics 55, 2005, pp 119-130 Galeotti M., Lanza A.,(1999) " Richer and cleaner? A study on carbon dioxide emissions in developing countries" Energy Policy 27, 1999, pp 565-573 10 Cole M.A.,(2004) " Trade, the pollution haven hypothesis and the environmental Kuznets curve: examining the linkages" Ecological Economics 48, 2004, pp 71-81 11 Lee C-C, Chiu Y-B., Sun C-H., "Does One Size Fit All? A Reexamination of the Environmental Kuznets Curve Using the Dynamic Panel Data Approach " Review of Agricultural Economics, vol 31, no 4,2009, pp 751-778 79 Socio-Economic Issues 12 Aslanidis N., Iranzo S., (2009) "Environment and development: Is there a Kuznets curve for C02 emissions? " Applied Economics 41, 2009, pp 803-810 13 Schmidheiny K., Basel, U (2011), "Panel Data: Fixed and Random Effects", pp 38-46 14 Gujarati, D (2003), "Basic Econometrics," 4th ed., New York: McGraw Hill, pp 638-640 15 Hsiao, C (2005), "Why Panel Data? Department of Economics Nanyang Technological University", Singapore and University of Southern Califomia Los Angeles, pp 279-382 16 Hausman, J.A., Taylor, W.E (1981), "Panel data and unobservable individual effects, " Econometiica 49, pp 1377-1398 17 Stem, D I., Common, M S., and Barbier, E B (1996), "Economic growth and environmental degradation: the environmental Kuznets curve and sustainable development." World Development 24, pp 1151 -1160 18 United Nation Conference on Trade and Development (UNCTAD), UNCTAD Statistics, http://unctadstat.unctad.org/ 19 Worid Bank (2011), "World Development Indicators DataBank 2011, 2012," http://databank.worldbank.org/ddD/home.do?Step=12&id=4&CNO=2 20 Yaffee, R (2003), "A primer for panel data analysis Connect: Information Technology at NYU", pp 27-35  ... countries become richer One of the 70 Environmental Kuznets curve for carbon dioxide emissions in ASEAN countries exceptions to this pattern is carbon dioxide emissions which seem to rise monotonically... of ASEAN countries for carbon dioxide emission Is roughly $14,000 per capita This result provides a projection that on average, ASEAN counfries may be able to reduce the levels of carbon dioxide. .. in carbon dioxide emissions The first reason is seen from a theoretical point of view: Unlike other air pollutants (such as nitrogen oxide or sulfur dioxide) which have local effects, carbon dioxide