Potential for Competition

This section considers LAC's 'potential for competition' with PRC in terms of exports to third markets, starting with the similarity of structures; the hypothesis is simply that the greater the similarity in export structures, the greater the potential threat from PRC – given its lower wages and faster expansion. Table 4 [ PDF 195.7KB | 1 page ] shows the distribution of regional exports by technology.

3.1 Technological Structure of Exports

In general when this type of breakdown is undertaken, most observers conclude that this export structure is unfavourable to the growth prospects of LAC, particularly relative to EA, as LAC has a much more limited focus on technologically sophisticated goods with dynamic market prospects.¹⁰ Within LAC, over the period Mexico (like PRC) shows a sharp decline in the share of primary and RB products. Mexico also behaves similarly to PRC in terms of the growing share of HT, but has a much lower share for LT products, counter-balanced by a higher MT share. The big 2, medium 4 and small 11 LAC economies all have high shares of primary and RB exports, with the larger economies having proportionately more MT exports. At the more disaggregated technology level, the highest reliance on mineral-based RB exports is in the Medium 4 (the impact of oil in Venezuela). Fashion cluster exports are relatively important for the Small 11, due to US outsourcing of apparel in the Caribbean and Central America (this will come under severe competitive threat from PRC after the end of the Multi Fibre Agreement at the end of 2004). MT process industries are significant for the Big 2 and the Medium 4, while auto products are most significant for Mexico and the Big 2. MT engineering exports are very significant in Mexico but not in other LAC economies; electronics are also large in Mexico and (because of Costa Rica) in the Small 11. Other HT exports are significant only in the Big 2.

PRC has a very different technological trade pattern from most of LAC. As noted, Mexico is only country that comes near it, but it still has significant differences. PRC has a much larger role for fashion cluster products and electronics, but a much smaller one for automotives. However, their growing similarity is particularly relevant, and we have noted the mounting Chinese threat to simple labour-intensive jobs in the Mexican maquilas.

These technology comparisons are fairly aggregate but they do suggest that Chinese exports do not pose a direct threat to the bulk of LAC exports, with some exceptions:

• Fashion products (of interest to the smaller economies and Mexico),
• 'Other LT' (this is a broad category but PRC may be posing a threat in specific products like toys, sports goods or travel goods that are exported by the smaller economies),
• Engineering products, where PRC is now a major exporter of machinery and consumer durables and may affect similar exports from Mexico and possibly Brazil. However their relative weight will raise transport costs and may reduce their competitiveness in markets to which LAC countries sell.
• Electronics, of export interest mainly to Mexico and Costa Rica.

Of course, these categories should be disaggregated to yield meaningful conclusions at the product and country level.¹¹

3.2 Product Structure

The similarity of export structures between LAC and PRC can be examined by product category (here we look at the 3 digit level for 181 products, excluding 'special transactions') without categorising them by technology. We start with the stability of export structures in each country, the correlation between export patterns in 1990 and 2002. A high coefficient shows that the export composition is relatively unchanging, while a low coefficient indicates structural change.

The more changeable structures are in PRC and Mexico (roughly correlation coefficients of 0.4 and 0.6 respectively) the least are the LAC Medium 4 and LAC without Mexico (correlation coefficients of over 0.9). It may be expected that more rapid structural change – if it allows the exporter to respond to shifting structures in world trade – will lead to faster growth. Thus is borne out by the data since a regression of the stability coefficients on export growth rates over 1990-2002 for our sample countries in LAC and PRC supports this expectation. The adjusted R-square is 0.31 (F=11.2) and the coefficient is negative and significant – 0.022 (t=–3.35). The high degree of stability in export structure in LAC, along with a specialisation in non-dynamic products, appears to be taking a toll in the growth of export earnings.

We now compare the export structures of individual LAC countries with PRC. For all products, Chinese exports overlap significantly only with Mexico and Costa Rica, and even here the correlation coefficient is relatively low (at only 0.47 and 0.27 respectively). Thereafter there is a huge drop in the coefficient, and all other LAC countries have almost no correlation with Chinese exports. As a comparison, PRC’s export structure and that of the main producers in EA has a correlation coefficient of 0.75 for 2002 (Lall and Albaladejo, 2004).

Taking manufactured products only, there is a fairly dramatic decline over time in the similarity of Chinese exports with most of LAC, due to the rapid structural shifts in the latter. Only Mexico and Costa Rica in 2002 have any significant similarity to PRC in 2002 (with correlation coefficients of around 0.5 and 0.35, respectively). Most other countries have correlations with PRC that are either negative or below 0.1. Even excluding RB products (where PRC is least specialized) from exports improves the correlation only slightly. Apart from Mexico and Costa Rica in 2002, now Jamaica and Colombia, but both in 1990, have coefficients of above 0.20. All the other countries, including Mexico in 1990, have lower coefficients. In terms of the current overlap, therefore, PRC seems to pose a very small threat to the bulk of LAC exports, including the large industrial producers of Argentina and Brazil: even excluding RB products their coefficients for 2002 are -0.1 and 0.13 respectively. Appendix Table 1 gives the export structure correlation coefficients for PRC and all the LAC countries.

3.3 'Sophistication' Structure

Another way of analysing export structure and similarity is the 'sophistication' of manufactured exports based on average income level of the exporter of each product: the higher the level the more sophisticated the product. 'Sophistication' captures technological and other product characteristics based on the location of export production: a product exported by richer countries has features that allow relatively high wage economies to compete and are (in the relevant period) out of reach of lower wage economies. For a given product, greater sophistication presumably embodies higher levels of processing and greater value added; the inability to raise sophistication with rising wages leads to the loss of competitive advantage.

As a simple comparison we calculate an average sophistication score for each country based on the scores of each of its products.¹² Table 5 [ PDF 83.9KB | 2 page ] shows the score for 1990 and 2000 for PRC, some of the countries in LAC and EA, along with some developed and poorer countries for comparison.

The industrialized countries are, expectedly, at the top, with the US in the lead. Each has a decline in its sophistication score over the 1990s, reflecting the shift in exports to lower wage countries. In fact, most countries, including developing ones, see a decline in their scores for this reason. Note that Ireland, a relative newcomer to the industrial world with a strong specialisation in (MNC driven) electronics, comes much lower (after PRC) for this reason.

Mexico comes just after Singapore, with a higher score than the two larger NIEs (Taipei,China and Korea) because of its concentration on autos, which has a higher sophistication score than electronics (the production of autos remains more a privilege of rich countries than electronics). The larger LAC economies, Brazil, Mexico and Argentina are in fact considerably closer to the EA NIEs than they are to PRC by this indicator.

The most helpful role for the sophistication index is likely to be in distinguishing different types of products within broad categories. The technology categories in Table 4 [ PDF 195.7KB | 1 page ] for example may contain a range of products of differing quality, subject to different marketing strategies and undergoing different degrees of processing and technology development. In such instances the index provides a simple way of differentiating within these broad technology categories.

Table 5 [ PDF 83.9KB | 2 page ] gives the average sophistication score within some important product categories for PRC and the LAC groupings, with Mexico shown separately.

The sophistication scores capture several factors and do not, as they stand, point to direct competitive effects within these broad categories. However it is interesting to note that despite its lower score for its total exports and the important category of textiles and clothing, PRC has higher sophistication scores that LAC and Mexico for automobiles and metalworking machinery (where its score is very high), as well as for instruments and industrial chemicals. Its very high scores may be due to a division of the supply chain by MNCs, which concentrate relatively sophisticated activities (that those normally carried out in developed economies) in PRC. For the important category of electronics the sophistication scores for PRC, LAC and Mexico are similar. Hence for the two largest product categories in terms of export value, textiles and clothing and electronics, the sophistication scores for PRC are either below or broadly similar to the scores for its LAC trading partners.

The views expressed in this paper are the views of the authors and do not necessarily reflect the views or policies of the Asian Development Bank Institute (ADBI), the Asian Development Bank (ADB), its Board of Directors, or the governments they represent. ADBI does not guarantee the accuracy of the data included in this paper and accepts no responsibility for any consequences of their use. Terminology used may not necessarily be consistent with ADB official terms..

Comment(s)

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