The effect of exports on carbon dioxide emissions: Policy implications

Bosupeng, Mpho Article The effect of exports on carbon dioxide emissions: Policy implications International Journal of Management and Economics Provided in Cooperation with: SGH Warsaw School of Economics, Warsaw Suggested Citation: Bosupeng, Mpho (2016) : The effect of exports on carbon dioxide emissions: Policy implications, International Journal of Management and Economics, ISSN 2543-5361, De Gruyter Open, Warsaw, Vol. 51, Iss. 1, pp. 20-32, https://doi.org/10.1515/ijme-2016-0017 This Version is available at: https://hdl.handle.net/10419/309616 Standard-Nutzungsbedingungen: Die Dokumente auf EconStor dürfen zu eigenen wissenschaftlichen Zwecken und zum Privatgebrauch gespeichert und kopiert werden. Sie dürfen die Dokumente nicht für öffentliche oder kommerzielle Zwecke vervielfältigen, öffentlich ausstellen, öffentlich zugänglich machen, vertreiben oder anderweitig nutzen. Sofern die Verfasser die Dokumente unter Open-Content-Lizenzen (insbesondere CC-Lizenzen) zur Verfügung gestellt haben sollten, gelten abweichend von diesen Nutzungsbedingungen die in der dort genannten Lizenz gewährten Nutzungsrechte. Terms of use: Documents in EconStor may be saved and copied for your personal and scholarly purposes. You are not to copy documents for public or commercial purposes, to exhibit the documents publicly, to make them publicly available on the internet, or to distribute or otherwise use the documents in public. If the documents have been made available under an Open Content Licence (especially Creative Commons Licences), you may exercise further usage rights as specified in the indicated licence. https://creativecommons.org/licenses/by-nc-nd/3.0/ International Journal of Management and Economics No. 51, July–September 2016, pp.20–32; http://www.sgh.waw.pl/ijme/ Mpho Bosupeng1 University of Botswana, Gaborone, Botswana The Effect of Exports on Carbon Dioxide Emissions: Policy Implications Abstract The purpose of this study is toexplore long run affiliations between exports and carbon dioxide emissions. This paper examines thirty-seven countries over the period 1960 to2010 and uses the Toda and Yamamoto causality approach toinvestigate the direction of causal links. The results reveal that carbon dioxide emissions Granger cause exports inthe following economies: Bolivia, Canada, Costa Rica, Morocco, Austria and Ireland. Nonetheless, the reverse causality proved that exports Granger cause carbon dioxide emissions intwelve economies. Furthermore, the study registered bidirectional causal links between exports and carbon dioxide emissions inthe USA and Burkina Faso. We conclude that countries should consider exports market demand, energy consumption and economic growth intheir attempts toreduce carbon dioxide emissions. Keywords: exports, carbon dioxide emissions, Toda and Yamamoto causality, energy consumption JEL:O44 Introduction In national income accounting, exports play avital role indetermining national output. This has led economists tocoin economic growth hypotheses, namely: the growth-driven exports hypothesis and the export-led growth hypothesis. The concern is that many economies are driven by exports, which works against controlling emissions. Asian countries, particularly China and India, are often defined as export-led economies. The glitch with DOI: 10.1515/ijme-2016-0017 © 2016 Mpho Bosupeng. This is an open access article distributed under the Creative Commons Attribution-NonCommercial-NoDerivs license (http://creativecommons.org/licenses/by-nc-nd/3.0/). The Effect of Exports on Carbon Dioxide Emissions: Policy Implications 21 export-led growth economies is that they tend tobe substantial carbon dioxide emitters. Countries such as China and India face pressure tolimit carbon dioxide discharges. China proposed a40–45% reduction incarbon dioxide emissions by 2025. According toGregg etal. [2008], China is the largest emitter of carbon dioxide globally and India reportedly has ahigh emissions growth of carbon dioxide. As both China and India became major economic players, their high growth has had amajor impact on the environment. Carbon dioxide is notthe only gas that has detrimental effects. There are other gases liberated, such as sulphur dioxide, which is associated with respiratory malfunctions and acid rain. Individuals living near mining industries have also reported early blindness as aside effect of sulphur dioxide emissions. The harmful outcome of excess carbon dioxide is the greenhouse effect, as carbon dioxide traps the sun’s heat resulting inan upsurge inthe earth’s temperatures. Ecologists have predicted that inthe long run, this process can lead toadramatic change inthe world’s climate. Economies that rely on food exports (such as rice, maize, and cocoa) will be affected adversely by this alteration. China has been active insoliciting new technology tocombat emissions through emissions trading system, and Malaysia stepped forward by introducing green taxation tolimit emissions. Even though carbon levies have proved tobe effective, their ramifications on economic growth are notimpressive. Manufacturing industries are limited by carbon taxation, which often leads toadecline ineconomic growth, especially incountries such as China and India. Most existing studies have focused on economic growth, carbon dioxide emissions and energy consumption. Few studies have attempted toaddress the direct relationship between exports and carbon dioxide emissions. Even though this relationship has notbeen studied indepth, the long run association of the variables carries robust implications for energy consumption, emissions reduction and economic growth. Many economies face adilemma: either they reduce carbon dioxide emissions or divert their resources toeconomic growth. It is important toidentify the main drivers of an economy before attempting toreduce emissions. If an economy is driven by exports as the main source of income, then careful measures should be enforced toensure that economic growth is nothindered. Where, however, acountry does notproduce high-tech exports, income could be accrued from sources such as tourism, education fees, income tax or rent. In that case, it would be easier toenforce policies reducing carbon dioxide emissions –such as green taxation. Countries that export minerals –such as copper (e.g. Zambia) and diamonds (e.g. Botswana) –have tobe more cautious with the magnitude of emissions as well as practicality inthe use of minerals. It is therefore imperative tounderstand the relationship between exports and carbon dioxide emissions. Hypothetically, if acountry intends toreduce emissions, policy makers can enforce energy conservation regulations, which should lead tolow emissions. However, other macroeconomic variables such as unemployment may then rise. For instance, an administrator operating alocal coal plant may then need less workers tooperate energy supply projects. Mpho Bosupeng 22 This paper proposes that exports rely on energy, which results inhigh energy demand. Consequently, exports growth increases energy consumption, resulting inhigh carbon dioxide emissions. By using the transitive property inmathematics, we can infer that exports production drives carbon dioxide emissions. This paper contributes tothe literature by examining the long term relationship between exports and carbon dioxide emissions. The question is, why is the relationship between exports and carbon dioxide evaluated inthis study? We posit two main reasons: First, this relationship is important indetermining market response toassociated exports. The expectation is that if carbon dioxide emissions are induced by exports, it implies demand for manufactured products. This creates arise inenergy consumption leading tothe production of more carbon dioxide. Second, this relationship facilitates the evaluation of the prudent use of resources, based on the presumption that exports induce an increase incarbon dioxide emissions and demand for energy. Economies need todetermine proper consumption channels of fossil fuels toavoid complete depletion. This means they have toconsider exports production and rising carbon dioxide emissions without exhausting energy sources. An alternative will be tocarry out aproduction process that will produce more energy with fewer emissions. However, this situation will differ depending on acountry’s energy sources. The investigation applies the Toda and Yamamoto [1995] causality process toanalyse causal links between economic variables. The rest of this paper is structured as follows: first, we review the literature, followed by description of methodology and empirical test results. We then discuss our results and reach conclusions. Literature Review The existing literature generally focuses on energy consumption, economic growth and carbon dioxide emissions. Alshehry and Belloumi [2015] confirmed significant long run relationships between energy consumption, energy prices, carbon dioxide emissions and economic growth inSaudi Arabia. Their study relied on the Johansen multivariate cointegration technique. The results showed that energy consumption stimulated economic growth and carbon dioxide emissions. This means that controlling energy consumption will impinge on economic development and the magnitude of emissions. The authors suggested that macroeconomic policies intended toreduce energy consumption and carbon dioxide emissions many notadversely affect Saudi Arabia’s economic growth. However, Wang [2013] suggested that upsurges incarbon dioxide emissions are caused by output growth inthe USA and China. The results are reasonable, particularly for China. China’s exponential economic growth has been attributed toexports. In this method of economic growth, industries also expand resulting inmore carbon dioxide emissions. Today China is the largest emitter of carbon dioxide inthe world. According The Effect of Exports on Carbon Dioxide Emissions: Policy Implications 23 toWang [2013] economies can reduce their emissions by minimizing energy intensity. Fossil fuels usually have ahigh energy value but their emissions are robust. Economies therefore need todetermine an equilibrium function that maximizes output and minimizes emissions simultaneously. Similarly, Omri [2013] found that energy consumption is the main source of carbon dioxide emissions inMENA countries. In another study by Zhang and Cheng [2009], acausality effect from energy consumption tocarbon dioxide emissions was observed inChina. That study further showed that both carbon dioxide emissions and energy consumption had nosignificant impact on economic growth. The authors suggested that since economic growth is notinfluenced by energy consumption or carbon emissions, China can carry energy policies that reduce emissions without affecting economic growth. These results were welcomed by China because it is under intense pressure tocut emissions and continue economic growth. The issue is the relationship between economic variables, which is notfixed. Over time empirical analysis can show that economic growth is directly influenced by energy consumption and carbon dioxide emissions. If that is the case, policy makers need torevisit energy policies frequently. In the Turkish economy, carbon dioxide emissions had asignificant bearing on energy consumption [Soytas, Sari, 2009]. As with China, nosignificant relationship between output and emissions was observed. Loganathan etal. [2014] revealed bidirectional causal links between carbon tax and carbon dioxide emissions inMalaysia. The authors noted that economic growth induced an increase incarbon dioxide emissions. However, green taxation stimulated economic growth inthe long run. Zhixin and Ya [2011] also suggested that carbon tax could stimulate economic growth ineastern parts of China. It is important tonote that there are other drivers of energy consumption such as population density [Jafari etal., 2012]. In China, carbon dioxide emissions growth was also exacerbated by the needs of transport sector. Xu and Lin [2015] showed that between 1980 and 2012 carbon dioxide emissions inChina’s transport sector increased approximately 9.7 times, with an average annual growth rate of 7.4%. Exports production has been associated with high carbon dioxide emissions. Chang etal. [2013] examined the ramifications of energy exports and economic growth using the bias-corrected least square model for five Caucasus economies (Azerbaijan, Armenia, Georgia, Russia and Turkey). The study proved that economic growth is brought by high energy exports and globalization. Kahrl and Roland-Holst [2008] observed that exports are the largest source of energy demand growth inChina. To summarize the literature review, the main driver of carbon dioxide emissions is energy consumption as the majority of economies use fossil fuels for energy and exports production. Fossil fuels are also used togenerate electric power. Additionally, automobiles use fossil fuels such as gasoline during internal combustion, which adds tocarbon dioxide emissions. In China, energy consumption and carbon dioxide emissions were found tohave nosignificant effects on economic growth. The literature shows that exports generate an increase inenergy demand. Few studies have investigated whether exports Mpho Bosupeng 24 affect the magnitude of carbon dioxide emissions. This research aims todetermine the direct relationship between exports and carbon dioxide emissions over the period 1960 to2010 using the Toda and Yamamoto [1995] causality approach. We postulate that exports increase energy consumption (high demand), leading tohigh carbon dioxide emissions. Denoting exports as E t ; energy demand as D e ; energy consumption as Ce and carbon dioxide emissions as CO2. The relationship can be represented as:  E t ↑ →D e ↑→C e ↑→CO 2 ↑  1 ( ) (1) Logically, we can evaluate the direct relationship between exports and carbon dioxide emissions as represented by the following anticipated relation:  E t ↑ →CO 2 ↑  2 ( ) (2) In summary, the literature fails toaddress the direct relationship between exports and carbon dioxide emissions. This paper aims tocontribute tothe existing literature by validating causal links between exports and carbon dioxide emissions using the Toda and Yamamoto [1995] causality approach. Data Sources and Methods This paper surveys the relationship between income accrued from exports and carbon dioxide emissions. The data was obtained from aweb source named The Global Economy (http://www.theglobaleconomy.com/). The data source contains anumber of different macroeconomic variables. This study examines thirty-seven different economies over the period 1960 to2010. Exports were selected inthis study because they are important inexport-led growth economies and should notbe dissociated with carbon emissions from manufacturing industries. Consideration was given tonet exports; however that will necessitate reducing exports data with imports. Focusing on exports is crucial as they are directly linked with manufacturing industries that require energy, and generate emissions. Many economies, inconsequence, are concerned with rising carbon dioxide emissions from exports production. Another plausible variable is GDP Carbon dioxide emissions were measured intonnes (t)while income from exports was measured inbillions of US dollars. The actual data were converted tonatural logarithms toperform our analysis. It is imperative that the data set is examined for unit roots. The Augmented Dickey Fuller test [Dickey, Fuller, 1979] was selected totest for stationarity of the variables. The testing procedure of the ADF is derived from the following generalized model: The Effect of Exports on Carbon Dioxide Emissions: Policy Implications 25 ∆ y t = α + β t + γ y t−1 + δ ∆y t−1 + ! + δ p−1 ∆y t−p+1 + ε t , 3 ( ) (3) The model applied inthis study is:  ∴∆yt= α + β t+ γ yt−1+ i=1 k ∑ δ i∆yt−1+ ε t.  4 ( ) (4) The definition of terms is as follows: the regression constant is α and β is the coefficient of the time trend. Following Asemota and Bala [2011], ε t was defined as the white noise error term. Eviews 7 was used totest the stationarity of the series. Table 1 shows the results of the Augmented Dickey Fuller test. TABLE 1. Augmented Dickey Fuller (ADF) test results Country Carbon dioxide Exports Barbados –2.596114* –1.615445* Bolivia –2.152875* –1.530077* Canada –1.664362* –0.630576* Chile –2.029469* –4.250026 Colombia –2.330316* –2.092489* Costa Rica –2.613937* –1.456228* Ecuador –1.064043* –1.604317* Guatemala –1.906877* –1.920948* Honduras –2.377123* –2.147829* Mexico –0.491741* –1.088600* Trinidad & Tobago –4.531461 –1.940269* USA –2.813025* –1.558457* Uruguay –2.173077* –2.121801* Venezuela –3.751882* –2.227163* Benin –3.201790* –1.773505* Burkina Faso –3.445926* –2.248234* Gabon –2.593545* –1.607382* Ivory Coast –1.890845* –1.782691* Kenya –3.501770* –1.637232* Madagascar –3.228420* –3.254481* Mauritania –2.611152* –1.452321* Morocco –1.612630* –2.294836* Niger –3.501239* –1.812805* Mpho Bosupeng 26 Country Carbon dioxide Exports Nigeria –2.174025* –2.213271* Rep. Congo –2.737782* –2.070720* Senegal –5.463986 –1.904137* South Africa –1.237836* –2.122675* Australia –1.243064* –2.178340* Austria –2.591350* –1.628513* Belgium –2.626649* –1.847323* France –2.805379* –1.291667* Ireland –1.770027* –2.043268* Italy –4.622520 –0.689090* Spain –1.106530* –0.711161* Sweden –2.749000* –1.331897* Switzerland –3.874268* –1.150608* UK –3.056598* –1.728430* Note: The ADF test statistics are reported above. The critical values for exports and CO 2 are as follows: –[4.152511] is the critical value at 1% level; –[3.502373] is the critical value at 5% level and –[3.180699] is the critical value at 10% level. Superscripts (*) indicate statistical significance at 1%, 5%, and 10% critical levels respectively. The results are based on the model: ∆yt= α + β t+ γ yt−1+ i=1 k ∑ δ i∆yt−1+ ε t.  Eviews 7 was used tocompute the ADF unit root test. The null hypothesis for the test is “series x, has aunit root”. S o u r c e : own elaboration. The above results show that the data is non-stationary at different levels (1%, 5% and 10%). This is proved by ADF statistics, which are greater than the critical values. The following series were stationary: Trinidad and Tobago, Senegal, Italy and Chile. The Toda and Yamamoto [1995] Approach toGranger Causality The aim of this paper is toinvestigate the long run relationship between exports and carbon dioxide emissions. The expectation is that as income escalates from exports production, carbon dioxide emissions will also intensify. The Granger causality test [Granger, 1969] was notselected because notall data inthis study are non-stationary. Granger causality also has several limitations. First, if the variables under consideration are driven by acommon third process with different lags, there is apossibility of failing toreject the alternative hypothesis of Granger causality. In addition, Granger causality is often based on the assumption that causal relations are aresult of cointegration. The advantage of the Toda and Yamamoto [1995] approach is that the VAR’s formulatedinthe The Effect of Exports on Carbon Dioxide Emissions: Policy Implications 27 levels can be estimated even if the processes may be integrated or cointegrated of an arbitrary order. Wolde-Rufael [2005] observed that the Toda and Yamamoto [1995] approach fits astandard vector autoregressive model inthe levels of the variables. In consequence, this minimizes the risks associated with the likelihood of mistakenly identifying the order of integration of the series [Mavrotas, Kelly, 2001]. The literature has developed anumber of cointegration methods following the contributions of Saikkonen and Lütkepohl [2000a, 2000b]; Johansen and Juselius [1990]; Johansen [1988, 1991]; Granger [1981]; Granger and Weiss [1983]; Engle and Granger [198]); Granger and Engle [1985]; Stock [1987]; Phillips and Durlauf [1986]; Phillips and Park [1986]; Phillips and Ouilaris [1986]; Stock and Watson [1987]; Park [1990, 1992]; Phillips and Hansen [1990]; Hovarth and Watson [1995]; Saikkonen [1992] and Elliot [1998]. Most of these studies are based on the assumption that cointegrated variables will be attracted toeach other inthe long run. Toda and Yamamoto [1995] noted that if economic variables are notcointegrated then the VAR should be estimated infirst–order differences of the variables tovalidate the conventional asymptotic theory. In consequence, the Toda and Yamamoto [1995] approach is applicable even if the VAR may be stationary, integrated of an arbitrary order or cointegrated of an arbitrary order. This study applies the Toda and Yamamoto [1995] approach as discussed by Wolde-Rufael [2005]. The testing procedure starts by augmenting the correct VAR order k by the maximal order of integration dmax [Wolde-Rufael, 2005]. Following this, a(k + dmax)th order of the VAR is estimated and the coefficients of the last lagged d max vector are ignored [Caporale, Pittis, 1999; Rambaldi, Doran, 1996; Rambaldi, 1997; Zapata, Rambaldi, 1997]. Denote exports as LX and assign carbon dioxide emissions as LE. The VAR system of the variables can nowbe depicted as: L Xt= α 0+ i=1 k ∑ α 1iLXt−i+ j=k+1 dmax ∑ α 2jLXt−j+ i=1 k ∑∅1iLEt−i+ j=k+1 dmax ∑∅2jLEt−j+ λ 1t  5 ( ) (5) LEt= β 0+ i=1 k ∑ β 1iLEt−1+ j=k+1 dmax ∑ β 2jLEt−j+ i=1 k ∑ δ 1iLXt−i+ j=k+1 dmax ∑ δ 2jLXt−j+ λ 2t  Empirical Results Eviews 7 was used tocarry out the Toda and Yamamoto [1995] approach tocausality. The results show that carbon dioxide emissions have asignificant influence on exports inthe following economies: Bolivia, Canada, Costa Rica, Morocco, Austria and Ireland. These countries registered ρ-values less than the 5% critical level, suggesting that we have