Disclaimer: This work has been submitted by a student. This is not an example of the work produced by our essay writing service.
You can view samples of our professional work here.
Any opinions, findings, conclusions or recommendations expressed in this material are those of the authors and do not necessarily reflect the views of UK Essays.
Base on the analysis above, we found China is a large CO2 emitter. However, there are huge amount CO2 emissions embodied in China’s international trade especially exports. A large proportion of exports focused on the sectors with high CO2 emissions coefficients. The EU introduces carbon tariffs on borders to tax the CO2 embodiments of imported products from the countries which have not implemented domestic carbon tax. It has not been implemented, because most EU states do not support this proposal except France and Italy. (Dissou and Eyland, 2011, Simon, 2012)
In this section we employ Gravity equation to estimate the EU’s carbon tariff elasticity of trade flow in several scenarios. The scenarios are designed based on our estimated CO2 emissions.
Gravity equation is first used empirically in Tinbergen’s work in 1962. It has been improved by later studies. Nowadays, it is intensively adopted to analyze international trade flow with influencing factors, such as economic size and geographic distance between countries. Normally, the economic size is estimated by GDP. It is proven that the trade between 2 countries is directly proportional to their GDP, and is negative to the geographic distance between them. (Feenstra, 2003b, Anderson and Van Wincoop, 2003).
Some research which use gravity equation add tariff variable into the model (Flam and Nordström, 2010, Hayakawa, 2013). Gravity equation is also applied to exploit the links between trade flow and tariff. Caliendo and Parro (2012) improve typical gravity equation to estimate the impacts of NATTA’s tariff reduction on both exports and imports of NATFA’s members.
Few adopt gravity equation on environmental issues. The research most closely related to our study are Aichele (2013), Larch and Wanner (2013). Aichele found that increase in one USD of the carbon allowance price declines EU’s carbon emission by around10 percent. And this 10 percent carbon emissions will be reallocated to other countries. Larch and Wanner (2013) study carbon tariff through gravity equation and indicate that carbon tariff decrease welfare of most nations but affect more seriously on poorer countries with lower technology of “clean” production. Trade will reduce as well. While carbon emissions will just move to low-tech-level countries from high-tech-level countries and global emissions raise up by 0.5%.
In this paper, we estimate of the EU’s potential carbon tariff elasticity of export of China, which could be defined by changes in trade flow between China and the EU per unit change in carbon tariff . We use EU’s applied lowest tariff rates instead of carbon tariff rates due to the fact that it has not implemented and even does not have specific rules on proposal. For export, instead of China’s exports value, we use the EU’s imports value in unit of USD with the top 19 trading partners of the EU (Algeria, Australia, Brazil, Canada, China, Hong Kong (China), India, Japan, Korea Rep., Mexico, Norway, Russian Federation, Saudi Arabia, Singapore, South Africa, Switzerland, United Arab Emirates, Turkey, United States, (EuropeanCommission, 2013)). Because the tariff targets of carbon tariffs are the EU’s imports. To simplify the model, we focus on EU-15 (Austria, Belgium, Denmark, Finland, France, Germany, Greece, Ireland, Italy, Luxembourg, the Netherlands, Portugal, Spain, Sweden and the United Kingdom). Overall, we estimate an improved gravity equation with tariff and trade data on 285 pairs of countries in last 25 years from 1988 to 2013.
Some earlier research Robertson and Estevadeordal (2009), and Hayakawa (2013) start estimation with the basic form of gravity equation derived by Anderson in 1979 and Anderson and Van Wincoop (2003), .
Where Xij is total value of exports from country i to j; Yi and Yj are GDPs of country i and j, respectively; Yw is world GDP; Tij denotes trade cost from country i to j, which can be measured by 1 plus tariff equivalents. Pi and Pi are explained as trade resistances in country i and j in work of Anderson and Van Wincoop (2003); σ represents the elasticity of substitution among variables.
Taking logs in formula [3.1], we obtain the basic estimating equation:
The trade resistances Pi and Pj can be measured by the geographic distance between country i and j; common land borders of country i and j; and common languages of country i and j and etc. (Gómez and Milgram, 2010)
Based on Anderson and Van Wincoop (2003), we build our gravity equation like:
i denotes exporting country, top 19 trading partners of the EU;
j denotes importing country, EU-15 member states;
s represents sector, there are 16 sectors, see Table1;
t represents year, year could be 1988-2013;
Xij represents imports values from country i to country j;
GDP represents GDP;
gdp represents per capita GDP;
Tariff ij represents tariff levied on country j’s imports from country i, Tariff ij equals to 1+ tariff rate ij;
Distance represents geographic distance between country i and country j;
area represents land area of a country;
contig ij is dummy variable valued 1 if country i and country j share a common land border, 0 otherwise;
comlanguage ij is dummy variable valued 1 if country i and country j use common language including both official and ethic languages, 0 otherwise;
colony ij is dummy variable valued 1 if country i and country j have ever had links of colony, 0 otherwise;
col45 ij is dummy variable valued 1 if country i and country j have ever had the same colonial relationship after 1945, 0 otherwise;
FTA ij is dummy variable valued 1 if country i and country j sign a free trade agreement, 0 otherwise;
country i is country dummy variable valued 1 if it is country j, 0 otherwise.
country j is country dummy variable valued 1 if it is country j, 0 otherwise.
μ ijst is error term, including omitted variables influencing trade flow.
In order to estimate tariff elasticity of trade flow, we adopt Ordinary Least Squares (OLS) to estimate our gravity equation. We follow the classic way of dealing with data: we do logarithmic transformation of Tariff as the same as GDP, per capita GDP, area and distance.(Hayakawa, 2013, Robertson and Estevadeordal, 2009) Therefore, the estimated coefficient of “Tariff” can be explained as tariff elasticity of trade flow/tariff power. Basing on the theoretical gravity equation and earlier empirical research(Hayakawa, 2013, Robertson and Estevadeordal, 2009), we expect Tariff has negative signed coefficient, because, normally, tariff plays a role as barrier of international trade. Theoretically, we expect the positive sign for GDP, contig, comlanguage, colony, clo45 and FTA, but a negative sign for distance. For per capita GDP and land area, there is no certain conclusion.
The EU has not introduced the specific rules. Normally, the carbon tax in the EU focuses on energy products, like fuel, and carbon intensive industries. In this paper, we design several scenarios to estimate the impacts of carbon tariff on trade under different tariff schemes. We have found out chemi and metal are the largest carbon emitting sectors. mq is fuel and energy sector. All of these 3 sectors are carbon-intensive sectors in China with high carbon emission coefficients. Thus we design 7 scenarios below:
Scenario 1: levy tariff on CO2 emissions from overall trade flow;
Scenario 2: levy tariff on CO2 emissions from all sectoral trade flow;
Scenario 3: levy tariff only on CO2 emissions in carbon-intensive: mq, chemi and metal;
Scenario 4: levy tariff only on CO2 emissions in carbon-intensive and trade-intensive sectors sectors: chemi and metal;
Scenario 5: levy tariff only on CO2 emissions in sector mq;
Scenario 6: levy tariff only on CO2 emissions in sector chemi;
Scenario 7: levy tariff only on CO2 emissions in sector metal.
We use data of EU’s imports in USD from the top 19 trading partners of the EU (Algeria, Australia, Brazil, Canada, China, Hong Kong (China), India, Japan, Korea Rep., Mexico, Norway, Russian Federation, Saudi Arabia, Singapore, South Africa, Switzerland, United Arab Emirates, Turkey, United States). For importing countries, we focus on EU-15 (Austria, Belgium, Denmark, Finland, France, Germany, Greece, Ireland, Italy, Luxembourg, the Netherlands, Portugal, Spain, Sweden and the United Kingdom). Overall, we use tariff and trade data on 285 pairs of countries from 1988 to 2013
The source of tariff rate data is theUNCTAD Trade Analysis Information System(TRAINS); the source of trade data is the UNSD Commodity Trade (UN Comtrade) database. We draw these data of tariff rate and trade from World Integrated Trade Solution (WITS) which is most powerful software jointly developed by The World Bank, the United Nations Conference on Trade and Development (UNCTAD), International Trade Center, United Nations Statistical Division (UNSD) and the World Trade Organization (WTO). We integrate these data to ISIC.Rev.3 classification through WITS.
For tariff data, we use effectively applied tariff rates which is the lowest applied tariff rate. For example, if both Most Favored Nation (MFN) applied rate and preferential rate exist for a product, the effectively applied tariff rate will be the lower rate for this product. The effectively applied tariff rates are reported by commodities, we integrate them into the form of sectoral tariff rates through WITS. There are simple average tariff rate and weighted average tariff rates for each sector, we use weighted average tariff rates for estimation due to the big differences of trade quantities among commodities. We also found the earlier studies use weighted average tariff rates for estimation through gravity model, such as Hayakawa (2013). The original tariff rate data is ad-valorem tariff rate, e.g. the tariff rate data: 1, which means this product will be levied a tariff valued 1% of its price. In our model we use log (1 + tariff rate), so the tariff rate here should be 0.01 when original tariff rate data are reported as 1. Tariff data we used is from 1988 to 2013. During the period of 1988 to 2013, most members of EU-15 use uniform tariff rate, because they are in the EU custom union, but Sweden has different tariff rates in few years for some commodities.
For data of trade, we use gross imports data of the EU’s top 19 trading partners. We exclude the trade data with zero value in estimation.
We obtain data of GDP, per capita GDP and from database of the World Bank(WB), These data are valued in constant prices of the year 2005. Free Trade Agreement (FTA) data is found in website of European Commission (EC). The data of distance, area, comlanguage, contig, colony, col45 are extracted from gravity database (CPII). For data of comlanguage, we combined common official language and common ethnic language sourced from CPII, if both common official and ethnic language are reported as zero; we record 0 for comlanguage, 1otherwise. For distance, we use distance between capital cities of country pairs.
1.3.1 Data comparison and problem
We found a problem of our OLS estimation: the positive autocorrelation exists in our model through Durbin Watson test (The results of Durbin Watson test are shown at Table 11-15). Autocorrelation means the error terms in different time periods or different sections are correlated. The positive autocorrelation could make standard error of coefficients too small even though the estimated coefficients are still consistent. Thus it will lead to a larger t-statistic value to reject the null hypothesis when it should be accepted. The autocorrelation can be corrected by adjusting the standard error of coefficient through the Hansen method.(Wooldridge, 2012) We do not correct this problem cause of time limitation.
In earlier studies about tariff estimated through gravity model, the coefficient of tariff are estimate about -20.18 to -1.01 in Melchior et al. (2009); -9.49 to -0.31 in cross section estimation in Robertson and Estevadeordal (2009); – 4.24 to -2.81 by OLS estimation in Hayakawa (2013). Compared with these studies, our estimated coefficients of tariff around – 4.07 to – 1.36 are acceptable and reasonable.
The elasticity of tariff estimated in last part will be used to calculate the amount of reductions in trade and carbon emissions. IO model will be adopted again.
First, we need to know the rate of the EU’s carbon tariff. However, the price of the EU’s carbon tariff has not been introduced officially. Normally the price of carbon tariff is designed to adjust the prices of imported products into the tax burden as same as the domestic products have and also try to achieve the goal of reflecting environmental cost of CO2 emissions. For this purpose, we assume the price of potential carbon tariff of the EU is the same as the price of domestic carbon tax price in the EU. Although the EU has not had a identical carbon tax price and not all member states have such carbon tax, it is reported that the EU Commission has suggested a tax of 4 to 30 euro on per ton of carbon for businesses(Kanter, 2010). In this paper, we use this suggested price as the carbon tariff price. We assume that there are 3 tariff levels: 4 euro per ton of CO2 for low level, 17 euro per ton of CO2 for middle level, 30 euro per ton of CO2 for high level.(Zhang, 2010, Kanter, 2010)
First, we calculate the carbon tariff rate on CO2 emissions of output in 1 monetary unit.
Where, R is carbon ad-valorem tax defined by the carbon tariff price of CO2 emissions from output in 1 monetary unit. The r denotes carbon tariff price per ton of CO2 emissions, in this paper it should be 4/ 17/ 30 EUR per ton of CO2 emissions. Eex represents total CO2 emissions embodied in China’s exports to the EU in unit of Million tons, which has been calculated before. Xex is China’s exports to the EU in unit of monetary (we have assumed that the China’s exports to the EU equals to the EU’s imports from China.)
Table 16 shows the carbon tariff in form of ad-valorem tax under carbon price of 4-30 EUR/ ton CO2 in China in 2009 as an example. We found carbon-intensive sectors have the extremely high ad-valorem tax rates compared to non-carbon-intensive sectors. The carbon intensive sectors ele-gs-w, chemi, metal and mq reach the ad-valorem tax rates by 132.91%, 63.43%, 34.13%, 25.95% , respectively under high tariff level of 30EUR/ ton CO2 emissions.
Table 16 The carbon tariff in form of ad-valorem tax (R) in 2009
Notes: High tariff level : 30EUR/ Ton CO2 emissions; middle tariff level : 17EUR/ Ton CO2 emissions; High tariff level : 4EUR/ Ton CO2 emissions.
Source: estimated by author.
Through the gravity equation [3.3], we have obtained the elasticity of Tariff(1+tariff rate): α5. In equation [3.3], keep other variables constant except Tariff. We have equation [3.5]:
Based on equation [3.5], we could obtain the rate of trade reduction by rewrite [3.5] to equation [3.6]:
Where, Rr denotes reduction rate of China’s exports to the EU; X0 denotes the values of China’s exports with the EU before implementing carbon tariff; X1 denotes the values of China’s exports with the EU after implementing carbon tariff. After we obtain Rr, X1can be estimated out through equation [3.6].
Then we can obtain the changes in CO2 emissions embodied in China’s exports by inserting the amount of changes in exports into our IO model. Here, we assume that carbon tariff only influent China’s exports not imports so that the reduction of CO2 emission embodied in China’s exports equals to the changes in China’s total CO2 emissions in following analysis.
Cite This Work
To export a reference to this article please select a referencing stye below:
“Thank you UK Essays for your timely assistance. It has helped me to push forward with my thesis.”
Related ServicesView all
DMCA / Removal Request
If you are the original writer of this essay and no longer wish to have the essay published on the UK Essays website then please.