سیاست های تحقیق و توسعه، رشد درون زا و اثرات مقیاس

This paper constructs a scale-free endogenous growth model and studies the determinants of optimal R&D policy. The model combines two of the main approaches to removal of scale effects: the rent protection approach and the diminishing technological opportunities approach. The steady-state rate of innovation is a function of all of the model's parameters including the R&D subsidy/tax rate. Thus, growth is fully endogenous. Numerical simulations imply that it is optimal to tax R&D when innovations are of very small and very large magnitudes, and to subsidize R&D when innovations are of medium size. Under a wide range of empirically relevant calibrations, the subsidy rate turns out to be positive and fluctuates between 5% and 25%.

Endogenous growth theory came at a crossroads with the Jones critique in the mid 1990s. First generation endogenous growth models predicted that the long-run growth rate of an economy increases in the level of R&D inputs and thus larger economies should grow at higher rates.1 In two influential papers Jones, 1995a and Jones, 1995b refuted this scale effect prediction by examining the post-war time-series data from industrialized countries. In response, a second generation of endogenous growth models has emerged. This literature offers three main approaches to remove scale effects: (i) Diminishing Technological Opportunities (henceforth DTOs) put forward by Jones (1995b), Kortum (1997) and Segerstrom (1998); (ii) Rent Protection Activities (henceforth RPAs) proposed by Dinopoulos and Syropoulos (2007); (iii) Variety Expansion (henceforth VE) proposed by Aghion and Howitt (1998, Chapter 12), Dinopoulos and Thompson (1998), Howitt (1999), Peretto (1998) and Young (1998).2

This paper has constructed a scale-free quality-ladders endogenous growth model that combines two of the main approaches to removal of scale effects: the RPA approach (Dinopoulos and Syropoulos, 2007) and the DTO approach (Segerstrom, 1998). I find that the steady-state equilibrium growth rate is a function of all of the model's parameters including the R&D subsidy rate. Hence, the model implies fully-endogenous growth. The presence of RPAs augments the effective replacement rate faced by the incumbent firms. The magnitude of this augmentation is positively related to innovation-deterring elasticity, which is endogenously determined. The optimal R&D policy exhibits an n-shaped relationship with respect to innovation size. When innovations are of very small and very large magnitudes, the optimal policy is a R&D tax, and for medium-size innovations, the optimal policy is a R&D subsidy. Numerical simulations imply that competitive markets typically underinvest in R&D and thus the optimal R&D policy is a subsidy. The magnitude of the R&D subsidy lies between 5% and 25% for plausible parameter values.