Co-Author: Ken Itakura

In recent years we have witnessed increasing prominence of trade in intermediate goods and services. The fragmentation of global value chains (GVCs) has been motivated by sourcing intermediate inputs from more cost-efficient producers in order to enhance efficiency. In estimating welfare and sectoral effects of mega-regional trade agreements (MRTAs), such as the Trans-Pacific Partnership (TPP) and the Regional Comprehensive Economic Partnership (RCEP), it is necessary to construct a global dynamic computable general equilibrium (CGE) model that incorporates the GVC structure.

After more than five years of official meetings, 12 countries negotiating the TPP reached final agreement in October 2015. However, in January 2017 the United States withdrew from the agreement. In March 2018, the remaining 11 countries signed the Comprehensive and Progressive Agreement for Trans-Pacific Partnership (CPTPP). Separately, ongoing negotiations for the RCEP among 16 countries were speculated to accelerate.

The main objectives of our study, now in a preliminary stage, are twofold. First, we use a dynamic CGE model to estimate economic effects of alternative sequencings of MRTAs and to evaluate the extent of losses to the United States from its withdrawal from the TPP. We offer two plausible sequencings. The first is the implementation of the 16-member RCEP, followed by RCEP＋Taiwan and then a Free Trade Area of the Asia-Pacific (FTAAP). The second is the implementation of an 11-member CPTPP, followed by an enlarged CPTPP and then an FTAAP. In addition, a hypothetical scenario in which the TPP includes the United States is considered for comparison. Second, we use a dynamic CGE model that incorporates the GVC structure to estimate the welfare and sectoral output effects of the CPTPP and RCEP, and we compare the results with those obtained from a conventional dynamic CGE model.

The data used for the conventional dynamic CGE model are from Version 9 of the GDyn Data Base, the International Monetary Fund’s World Economic Outlook Database, and the United Nations’ World Population Prospects. For the model that incorporates the GVC structure, additional data are from the OECD’s inter-country input-output tables. We modified the import demand in the dynamic CGE model to reflect the GVC structure in which producers and consumers determine demands for domestically produced goods and imports from different trading partners simultaneously.

We designed the following three policy scenarios.*

We first constructed a 23-region, 29-sector dynamic CGE model without the GVC structure and conducted policy experiments. The results can be summarized as follows:

(i) For countries that are members of both the RCEP and CPTPP, the welfare gains under Scenario 1 (Asian track) and Scenario 2 (Trans-Pacific track) are similar. Figure 1 shows welfare changes under both Asian and Trans-Pacific tracks for Japan, China, and Vietnam, expressed in percentage changes in real income. Welfare gains in Japan and Vietnam are slightly higher under the Asian track than under the Trans-Pacific track until around 2030, but the difference becomes negligible by 2032. By contrast, China's welfare gains are significantly larger under the Asian track because it is not a member of the CPTPP.

(ii) Scenario 3 is included to compare the real GDP effects of the TPP with and without U.S. participation. As indicated in Figure 2, the United States loses out on a gain of $200 billion in real GDP by 2035 if it never becomes a member of the TPP/CPTPP, and its gain is reduced by $100 billion if its participation is delayed by five years.

We then constructed 17-region, 14-sector dynamic CGE models with and without the GVC structure and compared the results between the two models. The main results can be summarized as follows:

(i) The incorporation of the GVC structure affects the overall welfare results only modestly, and whether it increases or decreases welfare differs across countries. For example, for the year 2035, estimated welfare gains using the model with the GVC structure are slightly larger than those using the standard dynamic CGE model for Singapore, Brunei, Thailand, and Australia, but the opposite relation holds for Japan, Malaysia, and Vietnam.

(ii) With the GVC structure, in which producers and consumers decide on the allocation and sourcing of intermediate and final products, output changes become greater in many sectors, as shown in Table 1. For example, the percent increases in output become considerably larger for motor vehicles in Japan, petroleum and chemical products in Singapore, and machinery and electronic equipment in Vietnam. Not shown in the table, the percent reductions in output of textiles and apparel for the CPTPP members in the Western Hemisphere (Canada, Mexico, Chile and Peru) become significantly greater.

This work-in-progress has provided some preliminary results. We are currently updating the data and recalibrating the model. We hope to complete the final version by Spring 2021.

________________________

*In a revised version, India will be excluded from the RCEP. The implementation period of the RCEP will be assumed to be 2021-2030. TPP-16 is TPP-11 plus Indonesia, the Philippines, Thailand, Taiwan, and the United States.

_________________________

Co-author Ken Itakura is Professor of Economics at Nagoya City University, Nagoya, Japan.