Latin America. This paper looks at this challenge and its implications. It begins by asking: Does manufacturing still matter for Latin America? It argues that the region cannot a\ufb00ord to turn its back to a well-proven road to development. It then moves on to show that endowments, produc- tivity, scale and the government\u2019s role, all work together to make China a formidable competitor. The importance of this challenge is con\ufb01rmed by an analysis of the trade data, which suggests a small impact so far, but a disquieting trend.
China\u2019s emergence in world markets raises pointed questions about the future of Latin America and the Caribbean (LAC) in the world\u2019s division of labor. The once dominant view that the economic future of the region was in manufacturing has long been challenged by traditional trade theory and in practical terms by at least three generations of Asian Ti- gers (e.g., Japan, Korea, Malaysia). China and its \u2018\u2018unlimited supply of labor,\u2019\u2019 rapid produc- tivity growth, and highly interventionist state has brought the practical challenge to unprece- dented levels. This paper, using mainly descrip- tive production and trade statistics, looks at the nature of this challenge and its implications.
It is divided into \ufb01ve sections, including this introduction. Section2 revisits a time-old ques- tion: Does manufacturing really matter for LAC\u2019s development? It argues that even though geography and endowments are not favorable, the risks and limits involved in natural resource specialization, the size of manufacturing in the region and the opportunities for policy improvement, all suggest that yes, manufactur- ing still matters. Section3 takes on the compet- itive challenges posed by China to LAC\u2019s manufacturers. It shows that endowments, pro- ductivity, scale, and the role of government, all work together to make China a formidable
competitor. Section4, using correlations and standard shift-\ufb02ow analysis, looks at the trade impacts for LAC of China\u2019s growing presence in the world markets. The results point to limited but increasing market share losses for producers in the region, a trend that seems to con\ufb01rm the severity of the Chinese challenge. The last section concludes by discussing, in gen- eral terms, the (di\ufb03cult) options available to policymakers.
Whether an analysts rates China\u2019s emergence as a positive or negative shock to LAC depends to a great extent on his views about the importance of manufacturing for the growth prospects of the region. It seems, then, appro- priate to begin this paper by arguing that LAC cannot a\ufb00ord to turn its back to a well- proven road to prosperity.
\u00b4guez-Clare, Elio Londero, and Jorge Chami Batista and the anonymous referees for their comments. None of them though should be held responsible for the views expressed here.
published \u2018\u2018The Economic Development of Latin America and its Principal Problems,\u2019\u2019 the issue of whether manufacturing matters would have been settled. Yet, the debate in the region seems to go on fuelled, \ufb01rst, by LAC\u2019s poor industrial performance and second by the fact that economists have yet to agree on the relevance of the \u2018\u2018natural resource curse.\u2019\u2019
Among those that are, explicitly or implicitly, pessimistic about the future of manufacturing in the region, the main argument remains rooted in the logic of comparative advantages. Yet, they do not always agree on the con- sequences of this pessimism. For instance,De
ural-resource abundant countries such as Can- ada, Australia, Sweden, and Finland, are very enthusiastic about deepening the region\u2019s spe- cialization in natural resources, seen as a path- way to a \u2018\u2018knowledge economy.\u2019\u2019
ferent perspective. They share the manufactur- ing pessimism on the grounds of geography and endowment, but unlikeDe Ferrantiet al.
blessing: \u2018\u2018Natural-resource-rich communities invest their resources in land, permanent crops, and extractive equipment and very little in hu- man capital, which has a very low return on a co\ufb00ee plantation or the equivalent.\u2019\u2019 (p. 547).
They also dismiss the natural resource suc- cess stories as \u2018\u2018not completely meaningful\u2019\u2019 to Latin America,inter alia, because countries like Canada, Finland, and Sweden \u2018\u2018may have made a commitment to broad human capital accumu- lation for non-economic reasons prior to the period when the private return to human capi- tal exceeded the private return to physical cap- ital.\u2019\u2019 (p. 5).
Although in\ufb02uential, both strands of this manufacturing pessimism seem to stand on shaky grounds. The bulk of the empirical growth literature, at the very least, does not support the natural-resources enthusiasm of the \ufb01rst strand (see e.g.,Easterly & Levine,
diversi\ufb01cation. The well-known success cases of resource-based development strike more as exceptions than rule. In all of them, transition into manufacturing was a crucial element of their story, and they made this transition count- ing on a human capital base that preceded the
natural-resource development and in a global environment where fragmentation of the pro- duction processes was still incipient, hampered by transport costs and tari\ufb00 and non-tari\ufb00 bar- riers. In other words, they were in better condi- tions to explore linkages and diversify into manufacturing.
Sure, Chile of the 1990s is a \u2018\u2018domestic\u2019\u2019 suc- cess story. Yet, Chile\u2019s success (which, by the way, still has close to 40% of its exports linked to one single product\u2014copper) is dwarfed by the growth, diversi\ufb01cation, and technological sophistication of the more manufacturing-ori- ented East Asia (Mesquita-Moreira & Blyde,
matched by the problems of Peru, Bolivia, Ecuador, and Venezuela\u2019s, some of which bear clear symptoms of Dutch Disease.1
Likewise, the argument that LAC\u2019s manufac- turing is doomed by geography and endow- ments seems to overlook both important pieces of theory and facts. The \u2018\u2018new\u2019\u2019 theories of endogenous growth, for instance, emphasize the importance of innovation and learning, which, in turn, depend fundamentally on the accumulation of human capital (which, by the way, takes place in \u2018\u2018manufacturing\u2019\u2019 rather than in \u2018\u2018agriculture\u2019\u2019 in the canonical model). That is, they give pride of place to a type of endowment that might be in\ufb02uenced by the complex interplay of geography and institu- tions, but that also respond to policy. One way of reading these theories is to argue that accumulation of human capital and the ensuing process of learning and innovation can change a country\u2019s destiny beyond its geography and natural endowments.2
One cannot also ignore the fact that after more than a decade of trade liberalization, an average of approximately 15% of the region\u2019s GDP is still produced in the manufacturing sec- tor (seeFigure 1) and that countries such as Mexico and Brazil are signi\ufb01cant exporters of manufacturing goods. True, the manufacturing share of GDP has been declining rapidly; out- put and exports of manufactured goods are still dominated by \u2018\u2018mundane\u2019\u2019 resource and labor intensive goods or are concentrated in the labor-intensive links of the value chain; and the region has been having di\ufb03culties to increase its limited share of the world market, being thor- oughly outperformed by East Asia (seeLall,
Yet, there seems to be more to LAC\u2019s manu- facturing tribulations than endowments and geography. The declining share of the GDP
can be seen, at least in part, as an inexorable adjustment to the manufacturing overshooting produced by import substitution policies (ISP). It might also re\ufb02ect a fall in the relative price of manufacturing goodsvis-a` -vis services driven by trade liberalization. As shown in
gone though a drastic process of adjustment to the \u2018\u2018norm.\u2019\u20193 That is, they have achieved in the 1960s and the 1970s a level of industrial- ization, measured by the manufacturing share of GDP, higher than predicted by its per capita income. Trade liberalization, mostly in the mid- 1980s and 1990s, seems to have played a part in
Fixed effects regression
y = 2.281x \u2013 0.1348x2 \u2013 6.8564
R2 = 0.06
Source:WDI and Penn World Tables 6.1
This action might not be possible to undo. Are you sure you want to continue?