It is common to employ the GTAP model to provide a quantitative assessment of the economic impact of the RCEP. This is a computable general equilibrium (CGE) model developed for trade analysis by the GTAP [6] . The latest GTAP database is GTAP 9, published in 2015 and containing data on 57 industries in 140 countries and regions for three years (2004, 2007 and 2011). For our analysis, we use the standard static version of the GTAP model and the GTAP 9 database for 2011.

To facilitate comparison with other mega-FTAs (including the TPP, FTAAP, and JCKFTA), we combine the 129 countries and regions in GTAP 9 into 27 regions. We retain the original 57 industries in the database. Table 1 provides additional details on the specific regional combinations.

Our scenario assumes the complete removal of all import tariffs among the RCEP members. However, it is unlikely that the RCEP would remove all import tariffs across all sectors among the RCEP participants. In so doing, our scenario provides an upper bound of the possible economic impact of the RCEP.

Given the limitations of the available data, we focus only on GHG emissions as the form of environmental load. We employ the GTAP CO₂ emissions database and GTAP non-CO₂ emissions database to measure the impact of the RCEP on GHG emissions. These databases therefore enable us to measure not only CO₂ emissions, but also non-CO₂ emissions, including methane (CH₄), nitrous oxide (N₂O), and fluorinated GHGs (or F-gases) (namely, tetrafluorocarbon, hydrofluorocarbons, and sulfur hexafluoride).

The GTAP 9 CO₂ emissions database details emissions resulting only from the combustion of fossil fuels, with the levels of CO₂ emissions calculated by multiplying the amount of fuel consumed by emission coefficients [7] . We assume that these emission coefficients remain unchanged with the liberalization of trade, and that the levels of CO₂ emissions will therefore change in the same proportion as the levels of sectoral fuel consumption. Therefore, we calculate the post-RCEP levels of CO₂ emissions by multiplying the initial level of CO₂ emissions for each sector by the corresponding change in sectoral fuel consumption from the GTAP model results. For example, CO₂ emissions produced by coal use in the electricity sector is calculated by multiplying the initial levels of CO₂ emissions resulting from coal use in electricity sector by the change in coal use in electricity sector derived from the GTAP results.

The GTAP 9 non-CO₂ emissions database provides the emissions data for three major non-CO₂ gases (CH₄, N₂O, and F-gases) [8] . In this database, the levels for each type of non-CO₂ emissions are associated with output, endowment use, and input use by industry and private households [9] . We assume that the levels of non-CO₂ gases change in the same proportion as the corresponding GTAP variables. This assumption allows us to calculate the post-RCEP level of non-CO₂ gases by multiplying the initial level of non-CO₂ emissions by the corresponding sectoral changes derived from the GTAP results. For example, the paddy rice sector emits CH₄. We then calculate the post-RCEP level of CH₄ emissions from land use in the paddy rice sector by multiplying the initial CH₄ emissions by the change in land use for paddy rice derived from the GTAP results.

Table 2 details the impact of the RCEP on real GDP, total exports, and total imports. As shown, under full trade

a. Other ASEAN includes Myanmar and Timor-Leste. b. ROW1 includes rest of Asian economies. c. ROW2 includes rest of the world. d. nec means not elsewhere classified.

liberalization, the 16 participants are likely to experience a positive average impact of the RCEP on real GDP, total exports, and total imports, which increase by 0.19%, 2.95%, and 5.19%, respectively. In contrast, non- RCEP economies other than Russia are likely to experience a negative impact on real GDP. In percentage terms, the increase in Vietnam is the highest across GDP (0.91%), and the increases in Cambodia are the highest across total exports (8.96%) and total imports (10.31%).

By way of a comparison, reference [2] analyzed the economic effects of an ASEAN + 6 (Japan, China, Korea, Australia, New Zealand, and India) FTA (effectively the RCEP) using the GTAP model and the GTAP 6 database (corresponding to 2001). Reference [2] showed that the ASEAN + 6 FTA, under a full liberalization scenario, also had a positive impact on the real GDP, exports, and imports of the ASEAN + 6 countries.

a. Other ASEAN includes Myanmar and Timor-Leste. b. ROW1 includes rest of Asian economies. c. ROW2 includes rest of the world.

Table 3 provides details of the impact of the sectoral output changes in percentage terms. Throughout the following discussion, we focus on Japan, China, and India, as the three largest RCEP members (in terms of real GDP), Australia, given its greatest growth in non-CO₂ emissions, as shown later in Table 4, and the United States (US) and the European Union (EU) as key non-RCEP-member economies.

As shown in Table 3, with full trade liberalization among the RCEP participants, farm output (defined as the agricultural sectors from paddy rice to wool, silkworm cocoons) tends to decline in Japan, but increases somewhat in China and Australia. In percentage terms, the decline rate in output from the paddy rice sector (−29.51%) is the most significant decline rate of all the sectoral output changes in Japan. In contrast, the increase rate in output

a. nec means not elsewhere classified.

from the paddy rice sector is the second most significant increase rate of all sectoral output changes in China (1.55%) and the third most significant increase rate in Australia (32.90%). Electricity sector output, one of the largest sources of CO₂ emissions, increases in Japan, India, and decreases in China, Australia, the US, and the EU.

Regarding the ASEAN + 6 FTA, reference [2] found that Japan was likely to experience a decrease in output in the agricultural and food sectors, while China and Australia were likely to experience an increase. With respect to the JCKFTA, reference [10] showed that Japan was likely to experience a decrease in output in the paddy rice sector, whereas China was likely to experience an increase.

Table 4 details the impact of the RCEP on GHG emissions. All figures are in million metric tons (Mt) of CO₂ equivalent. We first focus on the results for the 16 RCEP members and the world. As shown, the RCEP is likely to increase the total amounts of GHG emissions in both the 16 RCEP members and the world, with total GHG emissions of the RCEP and the world increasing by 19.32 Mt CO₂ eq. (0.11%) and 25.10 Mt CO₂ eq. (0.06%), respectively. As for our main research question, these GHG emission impact results show that the RCEP is likely to “increase” the total amount of GHG emissions in the 16 RCEP members and the world.

In other findings, the total CO₂ emissions of the 16 RCEP members and the world will increase by 7.08 Mt CO₂ eq. (0.06%) and 10.38 Mt CO₂ eq. (0.04%), respectively, and the total non-CO₂ emissions of the 16 RCEP members and the world will increase by 12.24 Mt CO₂ eq. (0.23%) and 14.72 Mt CO₂ eq. (0.12%), respectively. Among the non-RCEP members, the total GHG emissions of the US will decrease by 0.23 Mt CO₂ eq. (less than 0.01%), while those of the EU will increase by 6.55 Mt CO₂ eq. (0.14%).

We next focus on the results for the individual economies. In terms of CO₂ emissions, Japan and Korea will experience the greatest and the third greatest increase of all the economies, with the sum of its increases in CO₂ emissions (5.59 Mt CO₂ eq. and 4.14 Mt CO₂ eq.) almost equaling the increase across all 16 RCEP members (7.08 Mt CO₂ eq.) and the world (10.38 Mt CO₂ eq.).

In terms of non-CO₂ emissions for the individual economies, Australia will experience the greatest increase (11.92 Mt CO₂ eq.), which is almost equal to the increase across all 16 RCEP members (12.24 Mt CO₂ eq.). In percentage terms, New Zealand has the highest rate of increase in non-CO₂ emissions (5.00%), while the Philippines has the largest rate of decrease (−3.45%).

Table 5 identifies the industrial sectors that contribute most in terms of CO₂, CH₄, and N₂O emissions. We selected the top five sectors by the absolute change in each emission. We omitted F-gases from the table because

a. Figures in parentheses are percentage deviations from the initial period. b. Other ASEAN includes Myanmar and Timor-Leste. c. ROW1 includes rest of Asian economies. d. ROW2 includes rest of the world.

relatively few sectors emit these GHGs. In terms of CO₂ emissions, non-agricultural sectors appear in the top five sectors. Four of six economies in Table 5 have an electricity sector.

In terms of work elsewhere, reference [11] assessed the impact of a regional trade agreement that decreased import tariffs among the six East Asian countries (Japan, Korea, China, Indonesia, Thailand, and Vietnam), also selecting the top five sectors according to the absolute change for each of the GHGs. They also showed that in all six of the countries they analyzed, the electricity sector was one of the major CO₂ emitters.

a. Figures in parentheses are percentage deviations from the initial period. b. nec means not elsewhere classified. c. PADHE is PubAdmin/Defence/Health/Education.

Regarding CH₄ emissions, we observe a significant increase in the volume of emissions in the cattle sector in Australia (8.68 Mt CO₂ eq.), whereas in percentage terms CH₄ emissions from the cattle sector decreased most significantly in Japan (−15.29%). These particular results mainly arise because of the large increase in output from the cattle sector in Australia and the large decrease in output from the cattle sector in Japan, as shown in Table 3. In Japan, the paddy rice sector also shows the significant decrease (−1.05 Mt CO₂ eq.). All six economies in Table 5 also have a cattle sector, while all six economies other than Australia also have a paddy rice sector. Reference [11] also found that in all six countries they analyzed, the paddy rice sector was one of the top five CH₄-emitting sectors.

As for N₂O emissions, we identify a significant increase in emissions from the cattle sector in Australia (5.50 Mt CO₂ eq.), while in percentage terms the decrease in these same emissions is the second largest in Japan (−15.21%). Once again, these results are due to the large increase in output from the cattle in Australia and the large decreases in output from the cattle sector in Japan, as shown in Table 3. Five economies in Table 5 other than India also include the cattle sector in their top five sectors.

For both CH₄ and N₂O emissions, three or more of the top five emitting sectors in each economy are farm sectors. Worldwide in 2005, while CO₂ emissions are concentrated in the energy sector, agriculture accounts for the largest share of non-CO₂ emissions [12] . In sum, our results show that the RCEP is likely to increase the total amount of GHG emissions for both the 16 RCEP members and the world.