Maastricht Chronicles
Sunday, October 31, 2004
The WTO, Vertical Specialization and Institutional Coevolution: Some initial theoretical speculation
Introduction - This is a Very, Very Early Draft for Comment
The last fifteen years have been a bruising ride for proponents of the standard neoclassical development policy recipe. A now large literature demonstrates that countries and regions that have most closely followed the now standard economic recipe set out by the Bretton Woods institutions have grown less quickly than they had under the previous import substitution industrialization model, less quickly than the developed countries, and less quickly than some countries that have experimented with heterodox development policies (Stiglitz 2002). Empirical studies have failed to show that liberalization strategies can sustain economic growth in the developing world. Even studies on trade liberalization, arguably the least controversial aspects of the standard recipe, have yielded surprisingly ambiguous results. (Rodriguez and Rodrik 2000, UNCTAD TDR 2002). After reviewing recent the growth performance of developing countries in recent decades, UNCTAD (2002) concludes that they "rule out an unequivocal causal link from trade to growth."
Interpreting the reasons for the failure of the standard recipe has become the stock and trade of development economics. In a prominent contribution to the debate, Chang (2002) argues that the historical experience of the now developed countries (NDCs) is fundamentally at odds with the prescriptions of the standard recipe. Documenting the widespread use of active Industrial, Trade and Technology (ITT) policies by the NDCs during their catch-up periods, Chang argues that the current emphasis on liberalization and the efficiency of market mechanisms is based on a misunderstanding of the historical processes that led to catch-up in the NDCs. Such policies, he argues, place obstacles in the way of developing countries seeking to implement the policy regimes that have been most effective in the past. The current orthodoxy amounts, in Chang's suggestive image, to "kicking away the ladder" - denying developing countries the opportunity to apply the types of policies that made the rich countries rich.
Chang's argumentation proceed deductively, from observations of successful catch-up strategies, past and present, to generalizations about the proper conduct of development policy. The analysis, however, has relatively little to say about the specific mechanisms whereby the standard recipe's policies impede development processes.
This dissertation explores one possible set of such mechanisms: those generated by the World Trade Organization trade regime. It will begin reinterpret the current World Trade Organization regime as a product of institutional coevolution in the context of a new technological regime. Drawing on recent research on new patterns of production organization by multinational corporations, it will explore the fit between new production patterns and the global rules of the trade game. The purpose is to identify the mechanisms whereby the new technological regime, and the institutional forms that have arisen to complement it, may serve as obstacles to effective development policies.
Ladder-kicking 101
Chang's argument can be summarized briefly: the standard neoclassical recipe for development is a case of "do as we say, not as we did." While NDCs now insist the market mechanism is sufficient to provide the incentives needed for development, Chang shows that, historically, active ITT policies played a major role in their catch-up strategies. With few exceptions (e.g. Switzerland) NDCs managed to catch up with industrial leaders through a heterodox policy mix including tariff protection for infant industries, export subsidies, technology transfer facilitation, investment subsidies, public research, and a host of other interventions.
Proponents of the standard recipe see this policy menu as wrong-headed from start to finish - so much so, in fact, that substantial portions of it have been banned or substantially restricted by subsequent rounds of multilateral trade negotiations. Chang argues that, in historical perspective, NDC fervor for a laissez faire development model is not at all surprising: while nearly all successful catch-up strategies have entailed use of active ITT policies, once catch-up has been achieved countries have typically shifted positions in favor of liberalization.
Chang shows that enthusiasm for free trade only spread in Great Britain once British firms had achieved technological dominance in a number of key markets, and did not spread to the US, Germany, and other NDCs until they had managed to catch-up with Britain. In this view, internationally competitive NDC firms perceive the world market as an opportunity, whereas less competitive firms in less developed countries tend to perceive international competition as a threat. The resistance in the developing world to a more integrated trade regime is just as "natural," from this perspective, as the enthusiasm for freer trade in the developed world.
Historical perspective in historical perspective
For all its historical focus, Chang's arguments can be criticized for a peculiar sort of ahistoricism: even if one establishes that activist ITT policies were at the center of appropriate development strategies in Europe and North America in the nineteenth century, it does not necessarily follow that such strategies are relevant to today's developing countries. Clearly, the technological and scientific context of production has shifted considerably over the last 200 years, with a number of technological regimes rising and then falling in succession.
Define TR.
The question, then, is whether Chang's insights are applicable given the currently technological regime.To explore this question, it becomes necessary to identify the main features of the new technological regime, and its implications for production organization and institutional structures. The term "globalization" has dominated discussions of the changing nature of technology and production in recent years, yet much debate remains on the specific characteristics of globalization, about what makes globalization truly different. At its most basic, the term refers to closer cross-border economic integration. While the dramatic increase in international trade and foreign direct investment in recent decades is commonly cited as evidence of globalization, skeptics can easily retort that levels of trade and cross-border investment were even higher at the end of the nineteenth century. (Hobsbawm)
From a technological regimes perspective, the question appears to be wrongly framed. Rather than focusing on the degree economic integration, the technological regimes framework focuses attention on the specific features of new patterns integration, on the "how?" question rather than the "how much?" question of globalization. For globalized production to constitute a new technological regime, it must be qualitatively different from the patterns of production that preceeded it: more of the same does not a new technological regime make.
There is, however, evidence that the new, globalized patterns of production organization are indeed qualitatively different from those of the past. A growing literature on multinational corporate strategies bears directly on this question. The key insight of this literature is that trade in intermediate inputs has grown much more quickly than trade in final products in recent years.
Cite a bunch of stats about growing trade in intermediate inputs here.
-Yeats (2001) finds that trade in inputs has grown much faster than trade in final goods, with intermediates counting for 30% of world trade in manufactures.
-Hummels, Ishii and Yi calculate that the increase in exports from Vertical Specialization accounted for a third of world export growth.
-According to the US BEA (2002) in 1999, 93% of exports by US parent firms to their foreign manufacturing affiliates were inputs for further processing.
--> From Hanson, Mataloni and Slaughter (2004)
A bewildering series of competing labels have been proposed in recent years to describe this phenomenon, a sure sign that the activity is growing faster than academic attempts to come to grips with it. Labels range from the clunky (intra-mediate trade -Antweiler and Trefler 1997, International Fragmentation of Production, Helg 2004) to the politically loaded (outsourcing) to the quirky (slicing up the value chain - Krugman 1996) to the psychedelic (kaleidoscope comparative advantage - Bhagwati and Dehejia 1994.) Feenstra (1998) describes the new technological regime as one of "integrated trade and disintegrated production." But the most descriptively useful label was proposed by Yi (2002), who describes the new pattern of production organization as Vertical Specialization.
In vertically specialized production, the ricardian logic of specialization is taken beyond its traditional domain of trade in finished goods and into the value chains for individual products. In a number of industries characterized by discrete, sequential stages in the production of finished goods (e.g. electronics), vertical specialization allows firms to take advantage of cross-country differences in factor prices by locating each production activity in its lowest-cost country. The textbook example here is the Maquiladora industries on the US-Mexico border, where capital and knowledge-intensive inputs are produced in the capital and knowledge-rich US, then exported to labour-rich Mexico for labour-intensive assembly and packaging, with the finished product then re-exported to the United States for final sale.
Vertical specialization constitutes the organizational manifestation of a new technological regime. The breakneck rate of growth in vertically specialized production in recent years has been enabled by a range of technological innovations. At the most basic level, improvements in shipping and transportation have brought down "natural" trade barriers enough to make vertically specialized production practicable. The ICT revolution has lowered communications costs sufficiently to overcome the logistics and coordination challenges poised by vertically specialized production. The internet has received particular attention in this regard. (Gereffi 2001)
Institutional Coevolution: Why the WTO?
Yet, as with every shift in technological regimes, technological change in and of itself is only half the story. The other half, concerning the way the institutional context changes and "coevolves" with the new regime, will be the focus of my dissertation. The new technological regime poses a series of international policy coordination challenges that could not be met with the institutional scaffolding that developed to sustain the old, vertically integrated production regime of the postwar boom years.
In particular, a vertically specialized regime depends on a comprehensive, liberalized multilateral world trade system for institutional viability. Research emphasizes the high sensitivity of vertically specialized production to tariff rates - an intuitively plausible concern, given that inputs in vertically specialized value chains cross borders repeatedly, magnifying the effects of tariffs (Yi 2002, Hanson et al. 2004, etc.). Low tariffs and non-tariff barriers, institutional predictability and liberal investment regimes are key institutional components of the new regime.
Cite a bunch of stats about own-price elasticities of demand for intermediate inputs - Yi, Hanson, Slaughter, Feenstra...
The World Trade Organization can be seen as an instance of institutional coevolution, where new institutional forms are introduced to exploit the full potential of the new technological regime. Whereas previous technological regime shifts had tended to call forth institutional innovations limited to the national level, the Vertical Specialization regime is international in nature. It is not surprising, then, that the new institutional forms that have arisen in tandem with the growth of vertical specialization are themselves international.
Observing the WTO Rule-making Process
In effect, the trade-related rules agreed through the multilateral negotiating framework at the WTO directly impact firms' ability to exploit the potential of vertically specialized production. While no one would deny this, vertical specialization researchers have tended to take the multilateral trade regime as exogenous to the firms engaged in vertically specialized production. Firms, in this model, take the transnational rules of the game as they find them, and adapt their corporate strategy to them.
Anecdotal evidence, however, suggests that reality is a good deal messier, with firms working actively to influence the negotiation process through a variety of channels.
MNCs do not, to be sure, negotiate multilateral trade rules directly. However, by their very nature, they are directly affected by the outcomes of such negotiations. They are certainly large enough to mobilize the resources to lobby the WTO system. And their influence over certain aspects of the current multilateral trade regime have been well documented, the influence of pharmaceutical companies in pressing for adoption of the TRIPs and TRIMs agreements being the most often-cited example.
In some areas of political economy, firms' influence over trade policy decisions has become part of the standard academic model. Grossman and Helpman's influential 1994 model explicitly links tariff levels with firm lobbying activities within a national context. An extensive literature has grown arisen to provide empirical backing for the model, with quite some success.
The question, ultimately, is whether the institutional coevolution of the WTO and new patterns of vertically specialized production act as a kicking away the ladder mechanism, generating obstacles to the implementation of effective development policies in the developing countries.
One can hypothesize a number of routes through which the WTO regime may generate kicking away the ladder effects. Chang argues that in placing a number of active ITT development policies "out of bounds," the WTO accords can be seen as the heirs of the "unequal treaties" forced on peripheral countries by NDCs at the end of the nineteenth and the beginning of the twentieth century.
Recent developments, however, caution against simplistic interpretations of the rule-making dynamics in the WTO system. The collapse of the Seattle and, especially, the Cancun ministerial conferences due to resistance from groups of developing countries to NDC negotiating priorities show that WTO rule-making is not (or is no longer) a matter of rubber-stamping the policy agendas of the developed countries and their MNCs. Already, pressure from developing countries has forced the WTO to strike competition, investment and government procurement policy from the Doha Round's negotiating remit. A successful characterization of the WTO rule-making system, then, must take into account the complex interactions between developed and developing countries in the negotiations.
What is needed, then, is a research methodology able to precisely characterize these complex interactions. We shall adopt the Advocacy Coalition Framework to systematize the study of the WTO policy process.
The last fifteen years have been a bruising ride for proponents of the standard neoclassical development policy recipe. A now large literature demonstrates that countries and regions that have most closely followed the now standard economic recipe set out by the Bretton Woods institutions have grown less quickly than they had under the previous import substitution industrialization model, less quickly than the developed countries, and less quickly than some countries that have experimented with heterodox development policies (Stiglitz 2002). Empirical studies have failed to show that liberalization strategies can sustain economic growth in the developing world. Even studies on trade liberalization, arguably the least controversial aspects of the standard recipe, have yielded surprisingly ambiguous results. (Rodriguez and Rodrik 2000, UNCTAD TDR 2002). After reviewing recent the growth performance of developing countries in recent decades, UNCTAD (2002) concludes that they "rule out an unequivocal causal link from trade to growth."
Interpreting the reasons for the failure of the standard recipe has become the stock and trade of development economics. In a prominent contribution to the debate, Chang (2002) argues that the historical experience of the now developed countries (NDCs) is fundamentally at odds with the prescriptions of the standard recipe. Documenting the widespread use of active Industrial, Trade and Technology (ITT) policies by the NDCs during their catch-up periods, Chang argues that the current emphasis on liberalization and the efficiency of market mechanisms is based on a misunderstanding of the historical processes that led to catch-up in the NDCs. Such policies, he argues, place obstacles in the way of developing countries seeking to implement the policy regimes that have been most effective in the past. The current orthodoxy amounts, in Chang's suggestive image, to "kicking away the ladder" - denying developing countries the opportunity to apply the types of policies that made the rich countries rich.
Chang's argumentation proceed deductively, from observations of successful catch-up strategies, past and present, to generalizations about the proper conduct of development policy. The analysis, however, has relatively little to say about the specific mechanisms whereby the standard recipe's policies impede development processes.
This dissertation explores one possible set of such mechanisms: those generated by the World Trade Organization trade regime. It will begin reinterpret the current World Trade Organization regime as a product of institutional coevolution in the context of a new technological regime. Drawing on recent research on new patterns of production organization by multinational corporations, it will explore the fit between new production patterns and the global rules of the trade game. The purpose is to identify the mechanisms whereby the new technological regime, and the institutional forms that have arisen to complement it, may serve as obstacles to effective development policies.
Ladder-kicking 101
Chang's argument can be summarized briefly: the standard neoclassical recipe for development is a case of "do as we say, not as we did." While NDCs now insist the market mechanism is sufficient to provide the incentives needed for development, Chang shows that, historically, active ITT policies played a major role in their catch-up strategies. With few exceptions (e.g. Switzerland) NDCs managed to catch up with industrial leaders through a heterodox policy mix including tariff protection for infant industries, export subsidies, technology transfer facilitation, investment subsidies, public research, and a host of other interventions.
Proponents of the standard recipe see this policy menu as wrong-headed from start to finish - so much so, in fact, that substantial portions of it have been banned or substantially restricted by subsequent rounds of multilateral trade negotiations. Chang argues that, in historical perspective, NDC fervor for a laissez faire development model is not at all surprising: while nearly all successful catch-up strategies have entailed use of active ITT policies, once catch-up has been achieved countries have typically shifted positions in favor of liberalization.
Chang shows that enthusiasm for free trade only spread in Great Britain once British firms had achieved technological dominance in a number of key markets, and did not spread to the US, Germany, and other NDCs until they had managed to catch-up with Britain. In this view, internationally competitive NDC firms perceive the world market as an opportunity, whereas less competitive firms in less developed countries tend to perceive international competition as a threat. The resistance in the developing world to a more integrated trade regime is just as "natural," from this perspective, as the enthusiasm for freer trade in the developed world.
Historical perspective in historical perspective
For all its historical focus, Chang's arguments can be criticized for a peculiar sort of ahistoricism: even if one establishes that activist ITT policies were at the center of appropriate development strategies in Europe and North America in the nineteenth century, it does not necessarily follow that such strategies are relevant to today's developing countries. Clearly, the technological and scientific context of production has shifted considerably over the last 200 years, with a number of technological regimes rising and then falling in succession.
Define TR.
The question, then, is whether Chang's insights are applicable given the currently technological regime.To explore this question, it becomes necessary to identify the main features of the new technological regime, and its implications for production organization and institutional structures. The term "globalization" has dominated discussions of the changing nature of technology and production in recent years, yet much debate remains on the specific characteristics of globalization, about what makes globalization truly different. At its most basic, the term refers to closer cross-border economic integration. While the dramatic increase in international trade and foreign direct investment in recent decades is commonly cited as evidence of globalization, skeptics can easily retort that levels of trade and cross-border investment were even higher at the end of the nineteenth century. (Hobsbawm)
From a technological regimes perspective, the question appears to be wrongly framed. Rather than focusing on the degree economic integration, the technological regimes framework focuses attention on the specific features of new patterns integration, on the "how?" question rather than the "how much?" question of globalization. For globalized production to constitute a new technological regime, it must be qualitatively different from the patterns of production that preceeded it: more of the same does not a new technological regime make.
There is, however, evidence that the new, globalized patterns of production organization are indeed qualitatively different from those of the past. A growing literature on multinational corporate strategies bears directly on this question. The key insight of this literature is that trade in intermediate inputs has grown much more quickly than trade in final products in recent years.
Cite a bunch of stats about growing trade in intermediate inputs here.
-Yeats (2001) finds that trade in inputs has grown much faster than trade in final goods, with intermediates counting for 30% of world trade in manufactures.
-Hummels, Ishii and Yi calculate that the increase in exports from Vertical Specialization accounted for a third of world export growth.
-According to the US BEA (2002) in 1999, 93% of exports by US parent firms to their foreign manufacturing affiliates were inputs for further processing.
--> From Hanson, Mataloni and Slaughter (2004)
A bewildering series of competing labels have been proposed in recent years to describe this phenomenon, a sure sign that the activity is growing faster than academic attempts to come to grips with it. Labels range from the clunky (intra-mediate trade -Antweiler and Trefler 1997, International Fragmentation of Production, Helg 2004) to the politically loaded (outsourcing) to the quirky (slicing up the value chain - Krugman 1996) to the psychedelic (kaleidoscope comparative advantage - Bhagwati and Dehejia 1994.) Feenstra (1998) describes the new technological regime as one of "integrated trade and disintegrated production." But the most descriptively useful label was proposed by Yi (2002), who describes the new pattern of production organization as Vertical Specialization.
In vertically specialized production, the ricardian logic of specialization is taken beyond its traditional domain of trade in finished goods and into the value chains for individual products. In a number of industries characterized by discrete, sequential stages in the production of finished goods (e.g. electronics), vertical specialization allows firms to take advantage of cross-country differences in factor prices by locating each production activity in its lowest-cost country. The textbook example here is the Maquiladora industries on the US-Mexico border, where capital and knowledge-intensive inputs are produced in the capital and knowledge-rich US, then exported to labour-rich Mexico for labour-intensive assembly and packaging, with the finished product then re-exported to the United States for final sale.
Vertical specialization constitutes the organizational manifestation of a new technological regime. The breakneck rate of growth in vertically specialized production in recent years has been enabled by a range of technological innovations. At the most basic level, improvements in shipping and transportation have brought down "natural" trade barriers enough to make vertically specialized production practicable. The ICT revolution has lowered communications costs sufficiently to overcome the logistics and coordination challenges poised by vertically specialized production. The internet has received particular attention in this regard. (Gereffi 2001)
Institutional Coevolution: Why the WTO?
Yet, as with every shift in technological regimes, technological change in and of itself is only half the story. The other half, concerning the way the institutional context changes and "coevolves" with the new regime, will be the focus of my dissertation. The new technological regime poses a series of international policy coordination challenges that could not be met with the institutional scaffolding that developed to sustain the old, vertically integrated production regime of the postwar boom years.
In particular, a vertically specialized regime depends on a comprehensive, liberalized multilateral world trade system for institutional viability. Research emphasizes the high sensitivity of vertically specialized production to tariff rates - an intuitively plausible concern, given that inputs in vertically specialized value chains cross borders repeatedly, magnifying the effects of tariffs (Yi 2002, Hanson et al. 2004, etc.). Low tariffs and non-tariff barriers, institutional predictability and liberal investment regimes are key institutional components of the new regime.
Cite a bunch of stats about own-price elasticities of demand for intermediate inputs - Yi, Hanson, Slaughter, Feenstra...
The World Trade Organization can be seen as an instance of institutional coevolution, where new institutional forms are introduced to exploit the full potential of the new technological regime. Whereas previous technological regime shifts had tended to call forth institutional innovations limited to the national level, the Vertical Specialization regime is international in nature. It is not surprising, then, that the new institutional forms that have arisen in tandem with the growth of vertical specialization are themselves international.
Observing the WTO Rule-making Process
In effect, the trade-related rules agreed through the multilateral negotiating framework at the WTO directly impact firms' ability to exploit the potential of vertically specialized production. While no one would deny this, vertical specialization researchers have tended to take the multilateral trade regime as exogenous to the firms engaged in vertically specialized production. Firms, in this model, take the transnational rules of the game as they find them, and adapt their corporate strategy to them.
Anecdotal evidence, however, suggests that reality is a good deal messier, with firms working actively to influence the negotiation process through a variety of channels.
MNCs do not, to be sure, negotiate multilateral trade rules directly. However, by their very nature, they are directly affected by the outcomes of such negotiations. They are certainly large enough to mobilize the resources to lobby the WTO system. And their influence over certain aspects of the current multilateral trade regime have been well documented, the influence of pharmaceutical companies in pressing for adoption of the TRIPs and TRIMs agreements being the most often-cited example.
In some areas of political economy, firms' influence over trade policy decisions has become part of the standard academic model. Grossman and Helpman's influential 1994 model explicitly links tariff levels with firm lobbying activities within a national context. An extensive literature has grown arisen to provide empirical backing for the model, with quite some success.
The question, ultimately, is whether the institutional coevolution of the WTO and new patterns of vertically specialized production act as a kicking away the ladder mechanism, generating obstacles to the implementation of effective development policies in the developing countries.
One can hypothesize a number of routes through which the WTO regime may generate kicking away the ladder effects. Chang argues that in placing a number of active ITT development policies "out of bounds," the WTO accords can be seen as the heirs of the "unequal treaties" forced on peripheral countries by NDCs at the end of the nineteenth and the beginning of the twentieth century.
Recent developments, however, caution against simplistic interpretations of the rule-making dynamics in the WTO system. The collapse of the Seattle and, especially, the Cancun ministerial conferences due to resistance from groups of developing countries to NDC negotiating priorities show that WTO rule-making is not (or is no longer) a matter of rubber-stamping the policy agendas of the developed countries and their MNCs. Already, pressure from developing countries has forced the WTO to strike competition, investment and government procurement policy from the Doha Round's negotiating remit. A successful characterization of the WTO rule-making system, then, must take into account the complex interactions between developed and developing countries in the negotiations.
What is needed, then, is a research methodology able to precisely characterize these complex interactions. We shall adopt the Advocacy Coalition Framework to systematize the study of the WTO policy process.
