Avoiding the trap: The dynamic interaction of North–South capital mobility and technology diffusion

Abstract

This paper analyzes a stylized model of international capital mobility and diffusion of embodied technologies from North to South. The South can fall behind in terms of technologies or get trapped in a situation in which it is unable to attract foreign capital and embodied technologies if it is too far away from the technology frontier and if its absorptive capacity is too low. The paper reconciles the view that technological catching up is stronger the larger the technology gap with the alternative view that technological catching up is strongest at a medium technology gap. The closer the South is to the technology frontier the more beneficial is a higher income share of foreign capital.

Keywords:

• technology diffusion
• technology transfer
• capital mobility
• FDI
• human capital
• absorptive capacity

JEL Classifications:

• F21
• O11
• O33
• O47

Acknowledgements

The author thanks Johannes Bröcker, an anonymous referee, Fabrizio Zilibotti, Wolfgang Buchholz, Sonja Peterson, Gernot Klepper, Daiju Narita, Zhanna Kapsalyamova and Manfred Wiebelt for very helpful comments, and Sorin Krammer for his support.

Notes

 1. Gerschenkron (1962) studies the phenomenon of economic catching up of countries that have fallen behind.

 2. Aghion (2007), for instance, applies the N&P approach to examine the effect of education on growth.

 3. A_s and A_n are time dependent variables. Time indices of variables are not shown explicitly. Time derivatives are denoted by dots.

 4. This is satisfied in a multiplicative specification that we will use in Section 4, but not in an additive specification.

 5. For more detailed calculations see N&P and our calculations in equations (6) to (10), which are a generalized form of the following basic calculations.

 6. A_s is given by equation (9) with θ _s set to zero.

 7. Our basic relation implies imperfect technology diffusion. Perfect technology diffusion, i.e. ϕ _s →∞, would make own innovation in the South superfluous. For a further discussion see Grossman and Helpman (1991).

 8. Here we write indices s, because the influence of human capital, infrastructure, R&D, etc, in the South on Southern innovation is independent of the corresponding Northern values.

 9. Note that the start value A_s (0) in general includes knowledge gained from technology diffusion as well as from own Southern innovation.

10. While, in the former analysis a higher rate of technological progress λ of the technological leader made technology adoption more effective, the influence of λ on the technologies used in the South is now ambiguous, because innovation in the South is assumed to grow with λ, too.

11. Own innovation in the South may or may not be present at any rate γ ≤ λ.

12. The following outcomes also hold when including transaction costs of capital movements. In this case there can be a constant difference between the marginal products in the steady state.

13. Note that in this model the transfer of capital is beneficial both for the North and the South, because in the initial situation the mobile capital earns a higher return in the South than in the North and the return on foreign direct investment is transferred back to the North. Otherwise, no capital would be transferred from North to South.

14. The variable A has a maximal value of one, and the exponent of the first ‘A term’ in parentheses in equation (21) is always smaller than the exponent of the second ‘A term’ in parentheses so that the whole expression never becomes negative.

15. For reasons of mathematical simplicity, we have also chosen the income share of domestic low-tech capital to be α. But what matters for the model results is the income share of foreign high-tech capital, not the income share of domestic low-tech capital. Hence, the result would also hold for exponents that differ between K_i and D_i .

16. Alternatively, capital could be accumulated via an inter-temporarily optimal choice of consumption à la Ramsey.