What is the Evidence on Trade Creation?

PRC has seen a major increase in its imports from its neighbors in the region in the last few years and its rapid growth has been widely identified as a key source of dynamism for these countries. For example, from 1995-2003 exports of precision instruments and electrical machinery (much of this parts and components) from its nine major neighboring trade partners grew by over 600%; exports of machinery, chemical products and transportation equipment grew by around 300%.⁸ This import growth was over a period of major change in trade policy in PRC in preparation for WTO accession. Many of the changes needed for WTO entry were introduced during the 1990's, so that the weighted average tariff on manufactures fell from 47% in 1992 to 13% in 2001. Under the WTO agreement it is due to be reduced further to 7% (expected by 2005) and the remaining non-tariff barriers are to be simplified and phased out (Martin et al 2004).

However establishing the link between this surge in imports form the region and the trade reform associated with WTO entry requires more than simply a description or projection of current trade patterns. A counterfactual non-reform scenario must be compared with a projected 'with reform' case. The conventional means of addressing this is to apply a form of computable general equilibrium (CGE) model that compares a baseline (pre-reform) case with scenarios based on one or more trade reform packages. Roland-Holst (2002) and (2003) applies a version of the well-known GTAP model to assess the impact of reform on trade and income for both the region and PRC itself.⁹

The model provides a direct comparison of with and without scenarios and its outcomes are driven by a combination of assumed macro growth rates, changes in import protection (that is the degree of trade reform) and demand and supply patterns in the countries covered. However, as a projection of the future it is best described as 'indicative'; that is a projection of what will happen if markets clear in the way models of this type assume. As CGE models of this type assume that all markets revert to equilibrium, in effect they imply instant adjustment as resources shift from previously protected activities, so there are no frictional underutilization problems arising from changes in trade policy. This is the perfectly competitive world in which 'competitiveness' is not an issue. This is not to imply that such models give results that have no meaning, but by ignoring transitional difficulties arguably they have an implicit bias in favor of the policy they are examining.¹⁰ Furthermore there is always the issue of whether non-tariff barriers are accounted for adequately in this type of exercise.

The major result of Roland-Holst (2002, 2003) that after WTO accession in the longer-term up to 2020, PRC will have a rising trade surplus with N. America and Europe, but a rising trade deficit with ASEAN and the neighboring region more generally. Broadly speaking PRC will export finished goods to the former markets and import foodstuffs, raw materials, parts and components and capital goods from the latter..¹¹ Table 6a [PDF 59kb | 1 page] and 6b [PDF 53kb | 1 page] illustrates the basic run of the model, which compares the baseline case (that is projections under the assumption of no policy change) with the PRC WTO accession scenario.

Table 6a [PDF 59kb | 1 page] shows for example that PRC's exports to ASEAN are 36% higher in 2020 as a result of WTO accession, whilst ASEAN’s exports to PRC are 28% higher. The respective percentage changes for the Republic of Korea and Taipei,China (the NIE here) are 43% and 32%. Table 6b [PDF 53kb | 1 page] gives the same results now focusing on the change in the bilateral trade balance between different groupings and PRC as a result of WTO accession. The NIE for example have a bilateral trade surplus with PRC in 2020 of US$ 34 billion as a result of WTO accession which explains roughly one-third of the their total projected surplus. For ASEAN the share explained by WTO accession is much smaller (presumably because trade barriers were lower prior to accession) at less than 10% (US$3 billion out of a surplus of US$ 41 billion).

These results can be extended by accepting WTO accession as a given and posing the question, what additional trade creation results from new regional arrangements such as PRC's joining the ASEAN free trade grouping, the Asian Free Trade Area (ASEAN plus PRC) or PRC, plus Japan and the Republic of Korea, joining ASEAN (ASEAN plus 3)? Tables 7a [PDF 38kb | 1 page] and 7b [PDF 47kb | 1 page] provide the answers in terms of percentage change in trade flows in 2020, now compared with the scenario of PRC’s WTO accession, rather than the original baseline.

When PRC in ASEAN is examined a strong growth in PRC exports to ASEAN is predicted (47% above the level with WTO accession alone). Import growth from ASEAN is only modest at 2%, presumably on the basis that barriers in PRC are treated as already very low after WTO accession. ASEAN significantly reduces imports from third countries, so there is an important trade diversion effect (for example US exports to ASEAN are 6% lower and Japanese exports are 10% lower). In the case of the wider group of ASEAN plus PRC, Japan and the Republic of Korea most effects are magnified, with PRC’s exports to the latter two countries rising strongly. Again however the exports to PRC grow only modestly relative to the predicted level under WTO accession (Japan's are 2% higher and ASEAN’s 4%). This once more is due to the fact WTO accession is taken to have offered easy market access to PRC to exporters from these economies. There are now also greater diversion effects for exports of non-members than in the more limited ASEAN plus PRC arrangement (US exports to ASEAN, for example fall by 9%).

Detailed information on particular sectors can also be derived from this model. If one considers the relatively inclusive regional trade grouping of ASEAN plus 3 as compared with the WTO accession scenario by 2020 PRC exports are higher in 9 out of the 18 sectors in the model, the vast majority of total export gains are in just two sectors Processed Food (US$40 bill in 1997 prices) and textiles (US$ 8.5 billion) (Roland-Holst 2003, table 3.9). A disaggregated look into import and export flows at the sector level arising from the ASEAN plus 3 scenario is also possible utilizing a simple measure 'intra-industry competitiveness', essentially net exports relative to total trade in the sector.¹² Table 8 [PDF 50kb | 1 page] gives this measure of bilateral trade flows by sector in 2020 for the scenario of PRC joining ASEAN plus 3.

The sectoral picture, which emerges is that in general under this scenario PRC is a net importer of primary products, foodstuffs and energy and a net exporter of manufactures. This pattern is replicated in its projected trade with ASEAN. In the important electronics category the IIC figure 0.02 indicates a small trade surplus of 2% of electronics trade (imports plus exports) between PRC and ASEAN. For trade with the Republic of Korea and Taipei,China in general there is a projected net deficit in manufactures with the important exceptions of Clothing, Processed Food and Motor Vehicles; here there is a heavy deficit in Electronics with the IIC of –0.42 indicating a trade deficit roughly 40% of total trade in electronics with these two countries. Trade with Japan is projected to be in surplus with the exception of the more capital and technology intensive sectors in Manufacturing and Construction.

Concern has been noted that closer trade links with PRC may push ASEAN economies down rather than up the ladder of comparative advantage into lower skill activities. Evidence from the same modeling work casts doubt on this. The IIC indicator can be adjusted to reflect differences in skilled to unskilled labor ratios between sectors, and this labor-adjusted version of the IIC can be used to classify sectors into 'import dependent', 'trade neutral' and 'export oriented'.¹³ If one considers changes over the late 1990's (1996-2000) in bilateral PRC-ASEAN trade on a skilled labor content basis there was a substantial shift of 16 percentage points towards greater export –orientation (which was much greater than if the unadjusted data are used). The implication is that over this period ASEAN was increasing its net exports to PRC in relatively more skill intensive activities.

Examination of trade flows alone does not indicate income or welfare changes (and may imply the 'mercantalism fallacy' that exports are good and imports are bad). The modeling exercise also incorporates income change estimates calculated as future discounted income streams with a consumption and savings component. The fullest statement of these estimates by the same author is in Lee et al (2004), which looks at a shorter period 2005 to 2015 and appears to use a slightly different model specification to the earlier work.¹⁴ Table 9 summarizes the income effects by 2015 for three different scenarios – PRC's unilateral removal of all remaining trade barriers (PRC UNI), PRC joining ASEAN and ASEAN plus 3. For 2015 the income change by country and region for these scenarios is given relative to the baseline (broadly the PRC WTO accession scenario). These estimates are given in two versions with (Table 9a [PDF 59kb | 1 page]) and without (Table 9b [PDF 49kb | 1 page]) agricultural liberalization.

As is predictable in this type of model, since adjustment costs are assumed away, the wider the spread of the area of free trade the larger are the benefits. Hence ASEAN plus 3 is the preferred arrangement in terms of income change for all countries, apart from the excluded trading partner Taipei,China. Unilateral removal of remaining tariffs by PRC is a superior alternative for it and the rest of the world than its entry into the limited free trade area of ASEAN, although the latter is a superior option for ASEAN countries. If agricultural trade is excluded from the reform process, benefits to all parties fall and the Republic of Korea and Taipei, China can lose from PRC's unilateral trade liberalization.

As note earlier these modeling exercises mask complex internal shifts in resource allocation within partner economies as trade barriers are reduced. In PRC this will entail potentially complex shifts within agriculture (for example in relation to grains) and in parts of manufacturing (particularly in heavy industry, parts of which are often said to be highly inefficient). These modeling exercises imply that there is ample income growth to compensate potential losers and ensure a ‘Pareto optimal’ outcome. However with rising inequality and a fiscally constrained state, compensation is likely to be potential rather than actual and the adjustment process will almost certainly imply winners and losers.¹⁵ Similar points can be made concerning adjustments in partner economies.

There has been considerable concern in many countries including PRC that national domestically owned firms may be too small to compete in global markets. During the 1990's official policy in PRC identified a 'national team' of 120 large enterprises to be 'championed' although for range of reasons, including restrictions on mergers and acquisitions due to intervention by provincial authorities and what was seen as forced diversification, the 'national global giants' strategy has been judged a failure (Nolan 2001: 187). As yet there is little evidence from the trade data that this has been a serious hindrance for the economy and that in key sectors local firms are too small to compete.

Download this Discussion Paper [ PDF 161.8KB| 19 pages ].

Comment(s)

There are [0] comment(s) for this entry. Post a comment.