New Evidence on Classical and Technological Convergence in Manufacturing

This paper applies the Dowrick and Rogers (2002) diffusion model to manufacturing sectors, identified by the technological content of their production, as according the taxonomy of Lall (2000b). The aim of the study is disentangling neoclassical from technological convergence, in a panel of 50 developed and developing countries, observed at 5-year intervals from 1980 to 2000. Employing the System Generalized Method of Moments estimator, supportive evidence for both type of convergence is found, in all sectors, in three distinct samples: the whole panel and the ones of advanced and laggard economies. Moreover, it is shown that convergence dynamics differ among sectors and countries. In particular, capital accumulation is the main convergence inner driver in developed countries, while technological transfer determines the result in developing ones. Then, in developing countries, capital stock seems to be relatively better allocated in traditional rather than modern sectors, while high tech sectors open the greatest opportunities for technical upgrading in both kinds of economies. Finally, absorptive capabilities are found to play a twofold role: they partially overcome the diminishing returns to capital and they enhance technological transfer. This latter result confirms the importance of capability building process.