تحرک سرمایه و تله توسعه نیافتگی

In a two-country model with threshold effects, we study the robustness of underdevelopment traps to capital mobility. It is shown that capital mobility can make a country cross the threshold and bifurcate toward sustained growth, subject to household expectations. Another result is, however, that a multiplicity of growth patterns such as long-run divergence and convergence clubs can co-exist with perfect capital mobility, hence with an identical rate of return across countries. In this model, efficient development policies are a combination of, first, liberalization—as capital mobility makes growth possible—and, second, government intervention coordinating expectations into self-fulfilling growth beliefs.

Despite the consensus that prevails in the professional development community, the importance of foreign investment for growth remains a disputed issue. On the one hand, the notion that foreign finance spurs development has indeed shaped policies and aid programs to a remarkable degree. For example, a large number of developing countries, generally with active donor support, have set up foreign investment promotion programs.1 Openness, particularly to foreign capital, has been a central prescription of the “Washington consensus”.2 An empirical study by Cohen (1994) did support the view that foreign finance helps recipient countries improve the efficiency of their production processes and therefore facilitates growth. A host of static, partial-equilibrium models have correspondingly provided theoretical arguments showing how foreign investment—particularly foreign direct investment—can influence the recipient country's productivity at the micro level.3 Still, the debate is far from closed. Critics of what Rodrik (2001)4 calls the ‘obsession with global integration’ forcefully point at the diversion of ‘human resources, administrative capabilities and political capital’ toward the donor-imposed imperative of opening up borders—and away from the more difficult task of developing domestic growth policies. In this view, foreign investment is undoubtedly useful, but only when supported by adequate, country-specific domestic policies for growth. The empirical research reported by Hanson (2001) confirms that foreign investment alone is not likely to modify the growth prospects of host countries. What is the contribution of general equilibrium models of endogenous growth to the debate? In fact, one would hard-pressed to find any theoretical justification in this literature for the belief that foreign investment helps growth. The research pioneered by Grossman and Helpman (1991) provides convincing arguments that opening up a stagnant economy can improve its growth prospects. But there are no intentional capital flows—no foreign investment—from rich to poor countries in this line of analysis. Convergence effects are due to international flows of R&D, as knowledge bears some features of a public good. Even in an extended concept of capital incorporating knowledge, flows of capital are involuntary side effects resulting from cross-border spillovers, and can thus hardly be interpreted as describing the profit-maximizing competitive flows of foreign investment observed in the real world. 5 In the neoclassical economy of Barro et al. (1995), intentional flows of private capital to poor economies do happen—at the expense of the analysis of long-run growth patterns. A seemingly unrelated branch of the growth literature has studied underdevelopment traps in closed economies. As recently surveyed in Hoff and Stiglitz (2001), several mechanisms have been modeled that give rise to coordination problems and the accompanying development traps/threshold effects. The celebrated “big push” analysis of Rosenstein-Rodan (1943) and its reassessment by Murphy et al. (1989) analyze demand spillovers: growth in the high-productivity sector leads to higher incomes, which in turn leads to high demand for the goods in that sector—hence creating a coordination problem. In Azariadis and Drazen (1990), an overlapping generation model is used to show that if the private return to human capital is increasing in its aggregate stock, then a multiplicity of steady state growth paths will exist, including a no-growth path. D'Autume and Michel (1993) and Zilibotti (1995) similarly use reduced-form convex–concave technologies to find that multiple equilibria exist above a certain threshold, while below that threshold the only feasible solution is a zero growth equilibrium.6 These models suggest that the solution out of underdevelopment traps could be government intervention, through subsidies to education and production, or some other form of investment coordination. But the potential role for foreign capital to act as a private “big push”—just what the professional development community tends to emphasize—has not been considered in this sort of general equilibrium settings. Yet a convex–concave technology, when placed in a multicountry setting, provides exactly what is required to reconcile the intentional capital flows of the neoclassical economy (and the real world's)—decreasing returns at low levels of capital—with the nondecreasing returns needed at higher levels to analyze the multiplicity of long-run growth behaviors also observed empirically.7 This paper is also related to the analyses of Rodriguez-Clare (1996), Markusen and Venables (1999)8 and Venables (1996), which uses both demand and cost spillovers to study the relationship between coordination traps and openness. An interesting result of this line of research is that openness can indeed help a poor country escape development traps. One key difference with the analysis proposed here, however, is that these models are restricted to partial equilibrium settings, with the home country taken as “small”. This precludes analyzing the influence of threshold effects on global world dynamics. In this contribution, we construct a model where threshold effects are combined with capital mobility in a two-country economy. In autarchy, as noted, this sort of technology generates underdevelopment traps. But is the trap robust to capital mobility, or, to the opposite, can capital flows help the poor country cross the threshold? More generally, this two-country framework makes it possible to analyze the dynamics of the distribution of capital—and therefore GDPs—between two integrated regions when threshold effects are thought to be important. Will capital concentrate in one region, while the other stagnate? Or will the two grow or stagnate together? Our findings suggest that capital mobility does open up new possibilities when combined with threshold effects. Even though the two countries are perfectly integrated, with differentials in rates of return arbitraged away, both long-run “symmetric” and “asymmetric” equilibria are possible. In a symmetric equilibrium, GDPs will converge to one another and both countries will grow at the same rate. Interestingly, we will show that such long-run symmetric equilibria include the case of a stagnant country that bifurcates toward sustained growth after capital mobility is allowed. But long-run asymmetric equilibria will, in contrast, produce sustained GDP divergence, as capital is accumulated in one country only. The model therefore provides a simple theory in which “miracles”, “disasters”, bipolarization, long-run divergence and convergence clubs are open possibilities even though rates of return are identical across economies. Are there any policy implications from this analysis? The result from the model that a stagnant country can change courses through capital mobility appears to provide some support to the practitioners' view on foreign investment and growth. However, there is more. The growth pattern of the open economy is ultimately determined by household expectations. In this sense, development policies that are efficient in the model are a combination of, first, liberalization—as capital mobility will make growth possible—and, second, government intervention coordinating expectations into self-fulfilling growth beliefs. In this economy with threshold effects, liberalization and government intervention are complementary. Section 2 presents the assumptions of the model. 3 and 4 analyze asymmetric and symmetric equilibria. Section 5 discusses some welfare implications of the analysis. Section 6 reviews global dynamics and possible policy interpretations. Section 7 concludes.

This paper provides an analysis of how international capital flows can influence growth behaviors in the presence of threshold effects. One finding is that under mild conditions, this integrated economy displays multiple competitive equilibria that include sustained growth everywhere—even if the poor region, in autarchy, had been trapped under an underdevelopment threshold. The basic economic mechanism behind this result is that access to a larger integrated capital market makes it easier for the poor country to cross the threshold. The analysis thus provides some support to the practitioners' notion that competitive capital flows can help to lift a poor country out of stagnation. Yet this is not the whole story told by the model. The development policies that work in this setting combine external liberalization, here specifically capital mobility, and government intervention. This is because with threshold effects, liberalization of capital flows and government intervention work in a complementary way: while the former are useful to switch from a situation of poverty trap to a situation of multiple equilibria, the latter may still be needed to coordinate investment into a sustained growth equilibrium. The model also helps to explain why a wide variety of growth patterns can be observed in an increasingly integrated world economy. It provides a theoretical account of why economic integration and the accompanying convergence of rates of return are compatible with, and even help explain, rising international GDP cross-country inequalities and bipolarization. It shows that threshold effects can render asymmetric equilibria sustainable in the global economy. Specifically, if the poor country is sufficiently far off the richer economy when the two open up, expectations can drive the former to a “disastrous” situation, with domestic capital and income stagnating below the autarchic no-growth position.