Abstract
The approaching tenth year of the Doha Round with no achievements to celebrate indicates a failure of the World Trade Organization. Formal negotiations of the Round expired in 2005 without reaching a consensus, and informal negotiations were stalled in 2008. Thus, the Trans-Pacific Partnership (TPP), a recent initiative to deepen trade relations among countries bordering the Pacific, was greeted with applause and relief as a step in the right direction. This article discusses the region-wide Free Trade Agreement series of linked agreements that cover various members and issues. The recursive dynamic computable general equilibrium (CGE) model simulates two scenarios against the baseline, namely, a TPP agreement with China and without China. The preliminary results show that the TPP agreement without China cannot change the significant roles of markets and geography as the principal factors behind the economic integration of Southeast Asia with China. Trade and investment agreements facilitate market forces, they do not oppose them. The integration of the Asia–Pacific countries may benefit the US and other key economies.
INTRODUCTION
As part of the free trade area of Asia–Pacific, the Trans-Pacific Partnership (TPP) is recognised as an open and multilateral free trade agreement (FTA), which is being negotiated among the United States (USA), Australia (AUS), Brunei (BRN), Chile (CHL), Malaysia (MYS), New Zealand (NZL), Peru (PER), Singapore (SGP), Vietnam (VNM), Canada (CAN), Mexico (MEX) and Japan (JPN). China (CHN) was not invited to participate in this partnership because it is one of the biggest countries in Asia–Pacific region.
For most TPP countries, China is their largest export destination, and the consumer market with the highest potential. The intra-TPP member goods trade amounted to over USD2.7 trillion in 2012, while trade between China and TPP members was over USD1.4 trillion and accounted for over half of all TPP-member trade with other Asia–Pacific Economic Cooperation (APEC) members. Table 1 shows the share of bilateral merchandise exports/imports that account for total exports/imports between TPP countries and CHN. The share of exports to CHN of high-income TPP countries (AUS, NZL, JPN and SGP) accounted for one-fifth to one-fourth of their total exports. The share of exports to CHN of developing countries (VNM and MYS) accounted for nearly 10 per cent of their total exports. CHN plays a more significant role in imports. The average share of TPP member imports from CHN accounted for more than 17 per cent of their total imports; the share of imports from CHN for VNM, USA and JPN accounted for 23.8 per cent, 19.5 per cent and 22.1 per cent, respectively, of their total imports. CHN is the largest importer for these countries.
CHN considers the TPP framework a crucial component of US’ recent initiative to return to Asia, which poses economical and geopolitical challenges for CHN. CHN is actively pursuing its own FTA agenda to counterbalance this initiative by moving ahead with the CHN–South Korea and CHN–JPN–South Korea FTA negotiations and by constructing a regional web of FTAs. However, CHN has not completely closed its doors to the TPP. The country may join the TPP when its benefits outweigh its costs, but its effects on Chinese and other non-member countries have not been clearly defined as yet.
Most scholars claim that the theoretical influence of TPP on CHN may have a negative impact on its economy because some TPP countries are developing countries with exports similar to those of CHN’s, but sold at lower prices. Therefore, this partnership may trigger an export competition between CHN and these developing countries. The TPP membership of JPN may also be perilous to the economic interests of CHN (Song and Jinyuan, 2012). Li and Whalley (2012) used gravity equation-based methods to calculate the trade costs, which affirmed that the TPP initiative may hurt the Chinese economy. However, this negative effect is relatively small because of CHN’s geographical and commodity composition. This partnership will also adversely affect other non-member countries while serving member countries.
However, some scholars believe that CHN should not worry about the potential damage of the TPP. According to Huang and Feng (2011), TPP will not affect CHN significantly because Asia–Pacific countries rely on CHN economically. CHN has not directly joined the TPP, but 9 of 12 TPP countries entered agreements or negotiations with CHN to join the partnership. This multiple FTA system continuously progresses and compensates for the absence of TPP membership. Petri et al. (2011) compared two current trade agreements in Asia–Pacific, namely, the TPP and the Free Trade Area of the Asia–Pacific (FTAAP). They used the recursive dynamic CGE framework based on the neoclassical macroeconomic closure. They found that each agreement benefits all members, but such benefits remain relatively small until such agreements are expanded. The early stages of the TPP favour smaller, low-income economies such as VNM; the middle stages favour larger countries such as JPN; and the final stages favour the USA and CHN.
Share of Merchandise Trade Between CHN and TPP countries, 2010 (%)
Share of Merchandise Trade Between CHN and TPP countries, 2010 (%)
This article discusses how the TPP agreements with and without CHN could potentially affect CHN and other participating and non-participating countries if the proposal results in an FTA among participants. This paper also contributes to TPP literature in two ways. First, it establishes a non-competitive input-output table, which differentiates processing from non-processing trade in the CGE model. Firms in the processing trade enjoy duty-free imports of intermediate goods and capital equipment used for export processing. The agreement will more likely obtain an overestimated result, especially for MEX and CHN, if the processing effect in the CGE framework is not considered. Second, this article uses the recursive dynamic CGE model to simulate two scenarios against the baseline, namely, TPP agreements with CHN and without CHN.
The rest of the article is organised as follows. Section 2 describes the TPP initiative and analyses the Asia–Pacific industrial supply chain. Section 3 introduces the recursive dynamic CGE framework and our data. Section 4 provides the results. Section 5 concludes the paper.
Supply chain trade refers to the cross-border flow of goods, investment, services, knowledge and people associated with international production networks (Baldwin, 2013). Globalisation since the 1990s has been associated with a sharp decline in the share of high-performing nations in world income, manufacturing and exports. The big winners are a handful of industrialised developing nations who joined rather than built supply chains.
This article considers the 12 current TPP participants and the current members of the Asia–Pacific Economic Cooperation (APEC) forum as part of the Asia–Pacific region. This region is significant to the global economy because it comprises 40 per cent of the world population and contributes to nearly 60 per cent of global GDP. The Asia–Pacific region is significant as a burgeoning market and an integral part of global supply chains. The east and south Asian TPP members are strongly connected by the trade of intermediate goods and the involvement of the US in complex production networks across the Pacific. For example, 64 per cent of Asian non-fuel imports in 2009 were traded as intermediate goods, and over US$600 billion worth of intermediate goods were traded between Asia and North America (Williams, 2013).
Figure 1 plots a triangle trade relationship in manufactured goods between CHN and its major trading partners. Each sub-graph in Figure 1 contains countries that supply one or two intermediate manufactured goods, a country that supplies final manufactured goods, and countries that import and consume final goods. These graphs indicate that CHN has become an important supplier of intermediate manufactured goods for many neighbouring lower-wage countries such as VNM, South Asian less-developed countries (SSA), and East Asian less-developed countries (XEA), even if the country trades large quantities of final goods surplus with the US and has a large trade deficit in intermediate goods with other industrialised countries. These developing countries also trade surpluses of intermediate and/or final manufactured goods with the US. However, CHN remains the centre of final assembly for many manufactured products because some of its labourers in the final assembly activities are migrating to other low-cost countries.
Developing countries in Asia experience a similar trend of transfer of manufacturing industries from CHN. This trend behaves differently in different countries. For example, MYS mainly produces electrical and optical equipment (30–33, ISIC Rev.3.1), whereas the manufacturing industry that migrated from CHN to VNM mainly produces textile, leather and footwear (17–19, ISIC Rev 3.1). The TPP with and without CHN have different effects on these developing Asian countries.
MYS manufactures electronic equipment through the assembly process. The effect of TPP agreements with and without CHN on MYS this country is less different. The agreements with and without CHN do not affect the import prices of MYS electronic products in CHN. A TPP agreement with CHN promotes the export of Malaysian electronic products by opening the Chinese market. VNM manufactures textiles, leather and footwear through traditional labour. The effect of a TPP with CHN on this country is significant. This effect may possibly leave an impact on its trade by promoting manufacturing transfer from CHN and by expanding their exports to CHN. Therefore, different types of TPP agreement promote different types of integration in Asia–Pacific.
This article uses a multiregional CGE model similar to other widely used models that analyse the effects of trade and trade policies (Francois and Shiells, 1994; Shoven and Whalley, 1984, 1992). This model focuses on the real world economy and incorporates considerable detail on sectoral output and bilateral and global trade flows. It also links with a macroeconomic model to generate macro scenarios which enable the model to determine implied trade flows and sectoral structural adjustments for each region in a recursive dynamic framework that is consistent with the macro scenarios. The CGE model generates the pattern of production and trade for various economies and adjusts them to economic shocks specified in the alternative scenario to predict the direction of global economic growth.
Sequential Dynamic CGE Model
The CGE simulation model utilised in this paper comprises two parts. The first part is a comparative-static CGE model that simulates changes within a given year. The second part provides inter-temporal linkages and simulates changes between years.
Comparative-Static CGE Model
Our model focuses on the US and CHN and their top trading partners. We chose 25 regions and 41 production sectors from each region to represent the world economy. CHN and MEX have expanded processing zones which are modelled as separate economies, so 27 economies were generated in the model (Table 2).
Each economy has its own utility and production functions. These functions consist of 41 sectors that include goods and services producers, private households and government bodies. Each region engages in the trade of commodities and services with other regions in the model, subject to transportation costs. The prices of these commodities and services are determined by market clearing through international trade. The model regards economic agents as price takers. Their demand for commodities and primary factor services is based on cost-minimising and utility-maximising behaviours subject to production function and budget constraints. The production sectors manufacture a single commodity. The intermediate and final users of commodities and services classify these products by region of origin through the Armington specification (Armington, 1969). Our database implements the Armington specification at the producer and consumer levels, which are the source of imports determined at the agent level.
Regions and Sectors in the CGE Model
Regions and Sectors in the CGE Model
The final demand side of the model consists of households that purchase commodities and services while saving a portion of their income, which consists of returns to primary factors and net taxes. The output of the capital goods sector represents the aggregate investment in new capital goods in each region. The sum of household savings is equal to the sum of investment expenditures. Our model assumes that the preferences for welfare and household demands are separable, which considers the total utility as a function of sub-utilities with further sub-groupings. Regional welfare at the apex of the utility tree is derived from private household expenditures, government expenditures and savings. An economic mechanism that links federal government expenditures in US regions is not yet available. The simulations do not change the distribution of regional income across private and government expenditures and savings. The Cobb-Douglas function implements this assumption to describe the substitutions among the three components of welfare. This function also determines the household demands for composite commodities. The producing sectors in the model demand two types of input, namely, primary factors and intermediate inputs. The primary factor composite refers to the constant elasticity of substitution aggregate of land (where appropriate), labour and capital. The elasticity of substitution among primary factors depends on the industry. The primary factor composite and intermediate inputs have no elasticity of substitution (that is, a Leontief technology is assumed).
The model uses the Armington specification for international trade and determines the sourcing of imports at the producer and consumer levels instead of at the national level. However, two aspects of the model affect its interregional and international linkages. These aspects are drawn from the Global Trade Analysis Project (GTAP) model (Hertel, 1997). First, a global sector demands each regional transportation service sector to provide a composite service to ship commodities across regions. The relative contribution of each region to the global transportation sector does not change because of the simulation. Shipping services should be proportionate with the quantity of a commodity shipped along a particular route.
Second, the global sector mediates between regional savings and regional investment. The global sector has a portfolio of regional net investments offered to regional households to satisfy their demand for savings. The model examines the regional composition of net investment by assuming that a negative relationship exists between the expected regional rate of return on capital and the amount of investment in a region.
Each region has fixed endowments of land, skilled and unskilled labour, capital and other natural resources. Labour services and services from existing capital stock are assumed to be inter-sectoral and perfectly mobile, but specific to a region. This condition implies that capital and labour services are priced equally among all sectors in a region. These services are assumed to move freely between the export-processing zone and the rest of the Chinese and Mexican economies.
The comparative-static model simulates the changes within a given year (t). Our simulation framework incorporates physical capital accumulation for the economy as a whole to simulate the changes between years t and t+1. This equation is expressed as follows:
where Kr,t is the quantity of capital available for use in region r during year t, Ir,tis the quantity of new capital created in region r during year t, and Dr is the rate of capital depreciation in economy r, which is treated as a parameter.
The comparative-static model determines the level of new capital goods or investment Ir,t. We generated a baseline based on static expectations. This baseline describes the evolution of the world economy in the absence of change that we want to analyse. Our baseline runs from 2007 to 2018, and from the projections for land availability, labour, population and GDP growth rates. The model considers population, labour and land availability as exogenous variables. Our projections indicate that these exogenous variables experience shocks each year. The model considers GDP as an endogenous variable. We changed the closure of the model and adjusted an economy-wide technology parameter to target GDP. A policy line was generated by simulating a shock that we want to analyse. The policy simulations include the baseline population, labour and land shocks that generate the baseline simulations, shocks for the economy-wide technology parameter determined in the baseline simulations, and the shocks that we want to analyse. The shock that will be analysed for a particular variable, such as total Chinese exports, creates a distance between the policy line and the baseline. We simulated a policy shock that occurred in 2013, which is the same year when the policy line initially deviated from the baseline.
Data Used in the CGE Model
We constructed a single-year global Inter-Country Input–Output (ICIO) table based on the GTAP database Version 8 and the processing trade information from CHN and MEX to provide a workable dataset for the global value chain and TPP negotiation analysis. The bilateral trade flows in the GTAP database were initially allocated into intermediate and final uses using the United Nations (UN) Broad Economic Categories (BEC) method and detailed trade statistics with six-digit HS code. The expanded input–output (IO) table for CHN with separated accounts for processing exports was based on Koopman et al. (2008). The 2003 MEX IO table with separate domestic and maquiladora represents a Mexican statistical agency. We integrated the IO tables for CHN and MEX with the GTAP database Version 8 using a quadratic mathematical programming model to minimise the deviation between the resulting data set and the original GTAP data (Wang et al., 2010). The database contains 27 economies with 41 sectors that support the initial global CGE modelling and FTA analysis.
The benchmark equilibrium serves as a baseline for counterfactual simulation analysis. Our baseline runs from 2008 to 2018 and incorporates the projections of the Food and Agriculture Organization (FAO) for land availability and that of the International Monetary Fund (IMF) for labour, population and GDP growth rates. We used the projections of the International Labor Organization (ILO) for skilled and unskilled labour. Population, labour and land availability were considered exogenous variables in our CGE framework.
TPP negotiation establishes a free trade area and completely eliminates import tariffs among TPP countries. This negotiation also considers institutional area, technical investment, services and government purchase, which implies the reduction or even the removal of other non-tariff barriers. This paper analyses the effect of the complete elimination of tariff barriers on exports, imports, terms of trade, real income and the labour market, and analyses percentage changes before the financial crisis (GTAP version 8.0) to determine such effects.
Rules of Origin
The Rules of Origin (RoO) must be followed to determine the national source of a product. These criteria are important because duties and restrictions depend on the source of imports. The US is consistently pressed for protectionist RoO for its sensitive products. However, the East Asian economic integration is market-driven rather than institution-driven. Production networks continue to flourish despite FTAs, wherein the RoO complicate the sourcing of imported intermediate inputs. This article ignores the effect of RoO in the CGE framework for two reasons.
First, average tariff rates are relatively high in agricultural and textile industries, which are considered as ordinary trades rather than processing trades. The Most Favoured Nation (MFN) tariff must be considered because TPP countries are also members of the WTO. The average applied MFN tariffs significantly vary among TPP countries. For example, VNM has an average rate of almost 10 per cent, whereas SGP charges an average rate of 0 per cent on the few items that it has produced. The average tariff rates of products are below 10 per cent for TPP countries, but some industrial and agricultural sectors have relatively high tariffs. For example, the average applied MFN tariff rate on Canadian dairy products is 247 per cent, even if the overall Canadian average applied MFN tariff rate is less than 5 per cent. Table 3 shows product categories with the highest tariff rates for each TPP country. These products include dairy, clothing, beverages and tobacco, and sugar and belong to the ordinary trade pattern without RoO restrictions.
Highest Tariffs by Product Category, 2011
Highest Tariffs by Product Category, 2011
The second reason why this paper ignores the effect of RoO in the CGE framework is that the regional content of value-added is sufficiently high to conceal the effects of RoO to some extent. Figure 3 indicates that Asia–Pacific countries enjoy a higher degree of economic integration: the share of exports/imports between TPP negotiators and CHN (Figure 3(a)) accounts for over 50 per cent of their average total exports/imports. The degree of trade integration among TPP negotiators is nearly at 40 per cent, even when CHN is excluded (Figure 3(b)).
Trade Integration Between TPP Negotiators and CHN, 2010 (%)
Trade Implications
Figures 4 and 5 compare the real growth rates of exports between TPP members and non-TPP members on alternative tracks in 2013.
First, for TPP members, trade covered by the TPP with the CHN track is greater than that covered by the TPP without CHN (Figure 4), because TPP economies have relatively low barriers and most of their trade is already covered by FTAs, including the large trade flows within North America. Second, JPN, AUSNZ, USA, MYS and VNM are the major beneficiaries of the TPP with CHN, as the difference between the two tracks is more significant. The initial gains of these countries reflect their access to each other’s markets. Other Asian economies also gain at the early stage because of their agreements with China, Japan and Korea, which indirectly promote Asian trade integration. However, export changes are negative for some countries, such as the REA and EU 15.
In terms of goods trade, expansion of the TPP to include China and the other APEC members would encompass more TPP country trade than expanding the agreement in any other region including the EU. Nine TPP countries, including the US, also have agreements or are negotiating with the EU. TPP countries are also well connected to one another through their existing trade agreement. However, the impact of the TPP agreement on the European industrialised economies (EU15) is negative compared with the impact on European emerging economies (EU12) (Figure 5). One of the major reasons is that the structure of commodities trade between European industrialised economies and the US and Japan is competitive rather than complementary, especially in the automotive market. To strengthen commodities trade and investment in the Asia–Pacific area will undoubtedly threaten the exports of high-technology manufacturing products from the EU.
This situation mostly reflects the trade diversion under the preferential trade agreement, from efficient, non-participating exporters to participating countries that receive preferences. Trade diversion harms importers such as the Rest of the East Asian countries by forcing them to buy regional products at higher prices, and harms the exporter by deteriorating its trade (Lloyd and Maclaren, 2004).
Third, the effects of trade diversion under the TPP are mainly concentrated on CHN, but such effects are negligible compared with the size of the Chinese economy. These effects mirror the benefits gained by TPP countries, such as VNM and MYS, which compete with CHN for the US market. Section II explains that the effects of the two tracks are less different in MYS (Table 4) because firms in the processing trade already enjoy duty-free imports used in processing exports. By contrast, VNM expands its textile exports for the competitive price stated in the TPP agreement. The industry transfer from CHN to VNM in the TPP with CHN is faster than that under the TPP without CHN, because the former enjoys the lowest cost of transfer (Table 5).
Changes in Value Added and Exports in Electronic Products on Alternative Tracks for MYS
Changes in Value Added and Exports in Electronic Products on Alternative Tracks for MYS
Changes in Value Addition and Exports of Textile Products on Alternative Tracks for VNM
Different types of TPP agreements promote different types of integration in the Asia–Pacific region. From Figures 6 and 7 we see that the effects of TPP negotiations with and without CHN are less different in processing export-oriented countries such as MYS compared with ordinary export-oriented countries such as VNM. Thus FTAs will not be expected to change Asia’s existing production chain significantly in the short term.
Change in Industry Value-added on Alternative Tracks, 2013 (%)
Faster regional integration can produce more benefits, such as the creation of jobs. The increase in job opportunities would be about 2–3 for every 10,000 workers in the US and CAN. This number can increase to 10–60 for 10,000 workers in developing economies such as MYS, MEX and VNM. The increased use of labour computed from the demand side is attributed to two factors, namely, the expansion of production, which corresponds to the creation of jobs, and regional economic integration, which corresponds to the transfer of jobs. These results show that job transfer occurs between TPP negotiators and non-TPP negotiators because of increased productivity and deeper trade integration (Table 6).
Deviation of the Two Policy Lines from Baseline: Job Market (% change)
Deviation of the Two Policy Lines from Baseline: Job Market (% change)
Most TPP countries receive more benefits from the wider agreement. CHN also benefits from Asian regional integration, which provides welfare gains, supports the progress of political objectives and facilitates rebalancing. However, CHN receives more benefits than the US from an Asia–Pacific FTA. These two countries may be politically and economically competitive, but they both share powerful economic interests. Thus, some Chinese researchers proposed a pro-active Chinese stance toward the TPP, which involves active research and study and possible participation in the early stages of negotiations to influence their terms (Peter et al., 2011).
Among the major Asian economies, JPN substantially benefits from both tracks, but these benefits require JPN to overcome its political obstacles. VNM receives the most benefits from both TPP tracks. This unexpected result may be attributed to the high degree of overseas protection against its principal exports. VNM maintains strong trade with the US and JPN and a strong and competitive position in industries such as textile and footwear (whereas CHN is losing its competitive edge), high initial domestic protection and powerful scale effects in its principal production clusters. These factors boost VNM’s exports and terms of trade under the TPP agreement. India is not expected to participate in the agreement in the next five years because its economic and trade ties with European and the North American countries are deeper than its economic and trade ties with Asian countries, India may experience small trade diversion losses (around −0.2 per cent of gross exports in both tracks) and may not receive value-added chain benefits associated with liberalisation.
Over the past two decades, production has become internationally fragmented and specialised firms in different countries take part in the production process of a product but at different stages of the value-added chain. This globalised production system allows more in-depth specialisation and brings efficiency gains as countries specialise in the segments in which they have a comparative advantage. The dramatic increase of trade in goods belonging to the same industry but at different stages of production reflects the reorganisation of production on a worldwide basis. The Asia–Pacific region is significant not just as a burgeoning market, but also as an integral part of global value chains. The East Asian members, in particular, are highly connected through intermediate goods trade and involve the US in complex production networks spanning the Pacific, in which China plays a key role in the link. So the negotiations must bridge complex interests among countries and expand the FTA to bring greater benefits to members.
The TPP Agreement attracted worldwide attention from the public, media and political bodies because one-third of global trade occurs in APEC economies. However, the TPP negotiations do not include CHN because the country’s size, location and dynamism exert an inexorable gravitational pull that have made Southeast Asia its largest trading partner. This article discussed the region-wide FTA series of linked agreements that cover various members and issues.
Two scenarios, namely, the TPP agreement, with CHN and without CHN, were simulated against the baseline using the recursive dynamic computable general equilibrium model. Economic giants,the US and CHN were initially considered as the magnets for market integration. These countries attract smaller economies to both tracks, and view short-term benefits as modest, but the tracks create templates that may influence these giants to promote a wider regional integration in the future. Small countries, such as VNM and MYS, may receive significant, immediate benefits from these agreements, especially if they join the TPP agreement with CHN. The TPP agreement without CHN may not change the principal role of markets and geography in the economic integration of Southeast Asia with CHN. Trade and investment agreements facilitate rather than fight market forces. Compared with other narrow effects, an integration that spans the Asia–Pacific delivers greater benefits to the US and other key economies.