Andrés Álvarez (AA) & Carlos Andrés Brando (CAB): From around 1930 to the early 1980s most middle-sized and large Latin American economies underwent rapid industrial growth and transformation -entailing economy-wide structural change. However successful the processes of industrialization, the fact is that at the outset of the twenty-first century these industrial sectors had lost both dynamism and the widespread support theretofore provided by public policy in fields as diverse as trade and exchange, finance, taxation, and the targeted development of technology and infrastructure. How did this happen? Did industrialising projects in Latin America fail or were they dismantled?

Robert Kaufman (RK): Both. There is an ongoing debate about whether structural contradictions in ISI were responsible for its collapse -I am relatively agnostic on that issue. But it is relatively clear that fiscal shortfalls and public subsidies to deeply indebted state-owned enterprises led to what Jeffry Frieden termed “indebted industrialization” during the 1970s.¹ In my judgment, the crisis that exploded in Latin America in the early 1980s was rooted in these policies, particularly the accumulation of very large public sector deficits. Painful fiscal adjustments were unavoidable, and this became increasingly evident in the devastating hyperinflations that followed efforts to elide these adjustments with “heterodox shocks” in Argentina, Brazil, and Peru.

On the other hand, the fiscal adjustments that were undertaken in Latin America were made much more painful by the fact that the Reagan administration and the IMF (International Monetary Fund) insisted on maintaining the principle that external debt would eventually have to be paid in full. This was never realistic, but it had the effect of placing the burden of the negotiations with creditors almost entirely on the Latin American governments. The debt crisis, in fact, was not resolved until the “Brady plan” of the early 1990s implicitly accepted a write-down. More to the point of your question, a similar logic pertains to the debates over trade liberalization and privatization that began to emerge in the mid-1980s. Like the fiscal adjustments, it is difficult to dispute the idea that the existing “ISI model” was in need of “structural reform” at the onset of the 1980s. But the speed and scale of these reforms -along with the social disruptions they entailed- also owed a great deal to the economic pressure and ideological hegemony of the “neoliberals,” both in the global financial community and within Latin America itself.

Adolfo Meisel (AM): Well, it is true that we have had, in recent years, a process of de-industrialization. Industry, as a percent of GDP (Gross Domestic Product), has gone down, but I attribute this to a set of factors not all of which are necessarily bad or negative. Colombia has achieved a middle level of development, and we have started to experience the growth of new sectors of the economy because of changes in the demand patterns resulting from an increase in GDP per capita, and because of changes in our relative position in the international context. As income has gone up, the patterns of consumption among Colombians have changed. Increasingly, people are consuming new services such as education, leisure, travel, technology services, all of which have increased their share of GDP. This is a natural process of development and is considered a positive outcome.

There are other outcomes, one of which is structural. One reason for this de-industrialization, that I call structural, is the tremendous impact of China in global trade patterns. China´s economy, due to low wages, and also because of legislative conditions and regulatory stability, is able to produce at very competitive prices. As a consequence, many low-income countries in the periphery, such as the more advanced countries in Latin America, have felt China´s impact in their industrial competitiveness. Chinese products, for instance, are increasingly participating in the supply of industrial goods for the Colombian market. A third factor that affected the de-industrialization process in Latin America at the beginning of 21st century was the increase in the price of oil and, as a consequence, a Dutch Disease phenomenon that reduced the competitiveness of the industrial sector -this was nonetheless just a short-term shock. All of these three factors combined have resulted in our de-industrialization process.

AA & CAB: The experiences of various Latin American countries with industrialization have frequently been examined under the framework of an integrating model: import-substituting industrialization (ISI), State-led industrialization, Inward-looking development, etc. As recent literature findings suggest, however, the differences among nations essaying ISI (to adopt the most common label) remain striking. The timing varies, from early substitution in Argentina and Brazil in the late nineteenth-century to the late-late start of Peru in the 1960s. The underlying political economies vary as well, from competitive and democratic regimes such as Punto Fijo in Venezuela, to one-party rule in Mexico, and authoritarian-praetorian politics in Chile and Argentina. Finally, there is variation around the preferred mechanisms to implement the strategy. Does such diversity defeat any attempts at generalisation?

RK: There are certainly wide differences across Latin America in terms of the context and timing -and even intentionality- of the turn toward industrialization during the 20th century. Any serious analysis of the paths of development must necessarily take this diversity into account. Indeed, as Carlos Brando has persuasively argued, it is highly problematic to argue that Colombia ever systematically pursued an import-substituting strategy at all.² Similar questions might be raised about, say, Venezuela’s oil-dependent economy, or about industrialization within Uruguay’s very limited domestic market. All of this calls for careful analysis of the specific experiences of individual countries and caution in attempting to generalize cross-nationally.

Still, the analytical utility of generalization depends on what you are trying to explain. The concept of an integrating “ISI” model can, in fact, be useful as a ground of comparison for explaining differences among the more modernized countries of the region. The common experience of early industrialization in Argentina and Brazil, for example, can be a useful point of departure for examining the factors that led to such different outcomes in those two countries. An even broader base of comparison emerged with the region-wide spread of import-substituting ideas in the aftermath of World War II. Like wide cross-national differences in the take-up of neo-liberal ideas four decades later, the shared interest in import-substituting “models” across a significant number of Latin American countries offers an analytic opportunity to highlight how differences in factor endowments, political institutions, and power relations affected the way these ideas translated into economic policy.

Differences within Latin America do, it should be acknowledged, pose a significant challenge to cross-regional comparisons, particularly with the more dynamic “export-oriented industrialization” (EOI) models pursued in Korea and Taiwan after a briefer period of ISI. Gross distinctions between “Latin America” and “East Asia” are certain to founder on the internal heterogeneity of the two regions. I do think, however, that reference to the countries of East Asia can help to frame important questions about the reasons for “roads not taken” in Latin America. Was the failure to more vigorously promote industrial exports -a common feature of otherwise quite different Latin American economies- the result of differences in factor endowments, in the structure of the political system, in positions within the global economy, etc.? We will never have definitive answers to such questions, but they are still worth asking.

AM: The model of industrialization by import substitution in Latin America has general patterns that are shared by different countries, but it also has local characteristics. I will stress one local characteristic of Colombia: this country was not an extreme case, compared with some of the “horror stories”, of the import substitution industrialization. There was anti-rural bias in these policies in the sense that they produced regional concentration of income; also in the sense that these policies led to excessive protectionism and laziness of entrepreneurs who became accustomed to rent-seeking and corruption in obtaining licenses and subsidized credits. And I would reiterate that Colombia was not an extreme case of this. The reason for that, I believe, is the role of the coffee economy, by which I mean the political economy of coffee. Coffee, as it was produced in Colombia, had very good linkages, to use the framework devised by Albert ^{Hirschman (1970)}. Among those linkages was the fact that this was an export product where the producers were able to take a big part of the income. So, the gains in that sector were appropriated by the small producers, whereas in other countries the main export products were primary goods controlled by foreign companies or by the governments. In Colombia, coffee production was relatively spread out in terms of the owners of the productive capacity with, very good linkages, and the State did not have complete control over the process. Therefore, coffee growers, through various institutions, such as Federación Nacional de Cafeteros, were able to control some of the gains of import substitution industrialization that were observed in Latin America.

AA & CAB: The so-called ISI model implied a desire to transform the role played by Latin American countries in the international division of labour. The attempt to compete in international markets, not only with the developed economies but also with developing Asian economies, appeared (ex-post) a risky bet. To what extend did the shift in the political economy and the economic strategies of the Asian developing countries affect the possibilities of a successful insertion of Latin America into the world economy?

RK: A response to this question has to begin with the observation that the advocates of ISI envisioned a very different form of international insertion than did the EOI path which Korea and Taiwan began to pursue in the late 1950s. To state the obvious: the East Asian countries sought to link their economies more closely to the industrial economies of the core, initially by leveraging their supply of low-cost labour and then moving on to more capital-intensive forms of export-oriented industrial investment. The pursuit of an ISI model in Latin America sought, on the other hand, to reduce vulnerability to fluctuations in global commodity markets through the use of quotas and exchange rate policies that would reduce external competition in the manufacturing sector and encourage the import of necessary capital inputs. These differences, in turn, had implications for policies toward the skilling of the labour force and social security; the East Asian countries attached a far higher priority to education. To be sure, favourable trade treatment granted by the United States to manufactured imports from its East Asian allies might have helped to foreclose EOI options for Latin America, but the two models did not directly compete with each other in international markets. Rather, they appeared to evolve along parallel paths that did not intersect.

The question about competition with Asia becomes more relevant, perhaps, beginning in the late 1960s, when some countries (Brazil, Argentina, Colombia, and Mexico) sought to diversify into manufactured exports. Even then, however, the manufacturing sector remained strongly rooted in domestic markets, and the late entry into global markets put them at a disadvantage relative to developing countries in Asia. By that time, Korea and Taiwan had already moved up the supply chain, drawing heavily on earlier investments in human capital, while more labour-intensive manufactures were beginning to migrate to their poorer Asian neighbours. The more industrialized Latin American countries, in contrast, had neither of these comparative advantages.

The question of competition from Asia has become most relevant since the abandonment of ISI strategies and the turn toward more liberal trade policies. In particular, since the advent of NAFTA, Mexico’s place in the North American supply chain has faced ongoing competition from Chinese manufactures. But the economies in most of South America have evolved in a complementary direction, with China acting as a major market for commodity exports and as a source of investment capital.

AM: If we compare the experience of Latin America with the Asian Tigers, I think that we can conclude that the Latin American import substitution strategy was not as successful. I will emphasize that we should not underestimate the profoundly different historical and social conditions existing in the Asian Tigers compared to the Latin America case. If we compare, for example, with Korea, a very successful Asian tiger, the latter´s cultural conditions and its traditions, were completely different from those existing in Latin America. The latter countries, for historical reasons, including those which had probably the best chance of achieving very high levels of economic development like Argentina, were not able to be as successful as the Asian Tigers due basically to reasons related to political economy that were rooted in the history of the region. So, it is not simply, I think, a question of strategy.

AA & CAB: Whether the Debt Crisis that Latin America faced in the early 1980s is conceived as part of global economic forces that overwhelmed governments´ reactive responses, or whether it is seen as the corollary of inappropriate internal policies, which promoted inefficient industrialization, by the 1990s a shift towards a new economic paradigm was well under way. To what extent, the Debt Crisis served as a catalyst for the ideological consolidation of Washington-Consensus policies, irrespective of the virtues and achievements of the previous experience?

RK: As I stated above in my response to the first question, the debt crisis was to a significant extent home-grown, but was made more painful in many respects by how the IMF and the United States government structured negotiations with creditors. Starting in the mid-1980s, the crisis also played a decisive role in discrediting earlier development models and increasing pressure for the adoption of “structural adjustment” policies advocated by the World Bank and other international financial institutions.

I believe that, in fact, some movement in the direction of privatizing indebted state companies and toward establishing more open economies was necessary. But we will never fully resolve the fundamental question of whether the ISI model had actually been “exhausted” (or even what this means), or about how much change was needed to restart the Latin American economies. What is clear, though, is that -whatever the economic realities “on the ground”- both ideology and international power asymmetries played a major role in framing how the crisis was addressed.

Within Latin America, the crisis badly weakened the economic and political interests groups with stakes in the ISI model: industrial conglomerates, state enterprises, and public sector unions. Although the “neoliberal” alternative did receive support from some manufacturing and banking interests (for example, the Monterrey group in Mexico), market-oriented technocrats provided the primary impulse from their positions within the financial bureaucracies of the state.

Behind these technocrats, in turn, stood a range of international forces with deep commitments to a neoliberal vision. Ronald Reagan and Margaret Thatcher epitomized and promoted this vision during the 1980s, but it was shared and implemented by the public and private institutions that dominated the global system and controlled access to credit. Many officials within these institutions brought intellectual sophistication, knowledge of the region, and a degree of flexibility to the policy debates in Latin America; they were not fanatics and they could not unilaterally impose their views on their Latin American interlocutors. But they operated within an ideological frame and organizational structures that precluded systematic debate about the “virtues and achievements” of the previous system.

AM: In the case of Colombia, as in many other cases related to Colombia, the facts do not fit! Colombia did not suffer a debt crisis in the 1980s. We did not have a debt crisis, not because of any special virtue, but because we had a very strong coffee boom as a result of Brazilian crop frosts, and we experienced a different situation in that period compared to other countries in the region. So, Colombia did not have a debt crisis in 1980, although it did experience a shift in the orientation of its macroeconomic and foreign trade policies. I think this was also a result of international trends, and pressure from multilateral organizations. For example, in the Colombian case, the opening of the economy that took place in that period, was adopted slowly during the end of the 1980s and faster at the beginning of the 1990s. These policies were promoted, not publicly but privately, by the World Bank and the IMF. This was done through studies they endorsed, but also through indirect pressures. So, I think, these trends are not necessarily a response to the debt crises. Maybe the debt crises, in other countries, served as an excuse or as an additional argument for this policy shift, but this was not the case in Colombia.

AA & CAB: Did the shortcomings of coordinating institutions promoting the regional integration of the Latin American markets (such as the Andean Pact) have an effect on the failures of the industrialization process?

AM: The short answer is no. The problems of regional integration institutions in Latin America, such as the Andean Pact (Pacto Andino), are related to the tremendous instability of the majority of the region´s economies. That instability is what stood in the way of integration. The extreme example today is Venezuela, but Ecuador and Peru have also had problems of instability. The most stable country has been Colombia. So, I don’t think this is a problem of coordination. It is mainly a problem of the conditions of tremendous instability of these economies. Facing this instability, it is best for a more stable economy such as Colombia to search for integration directly to the world economy, as we have done through bilateral agreements.