A multi-sector growth model with technology diffusion and networks ∗ Manuela Magalh ˜ aes † Oscar Afonso ‡ August 30, 2016 Abstract This paper adds the production network into a multi-sector endogenous growth model to analyze the respective effects on technology adoption and thus on economic growth. In particular, we show that the higher the network degree, the higher is TFP and the stronger is the impact on extensive and intensive margins; hence, more likely is technology adoption and economic growth. Therefore, distinct cross-country network structures explain inter- country income differences. We then estimate the model and confirm theoretical findings: e.g., the network degree and technology invention year can together explain 81% of the technology adoption lag variation (extensive margin), the network degree can explain 9% of the intensive margin variation, and the inverse-network degree and the relative intensive and extensive margins together account for approximately 31% of the inter-country income differences. JEL classification: O33, O41, O47. Keywords: Income Differences, Production Network, Technology Diffusion. 1 Introduction The core of the endogenous growth theory should be not only the production but also the diffusion of knowledge, which is responsible for Total Factor Productivity (TFP) and explains the majority of inter-country differences in per capita output (e.g., Hall and Jones (1999); Jerz- manowski (2007)). However, since the initial seminal endogenous growth models (Romer, 1990; Grossman and Helpman, 1991; Aghion and Howitt, 1992) until the semi-endogenous growth models (Jones, 1995), the focus of the literature was only on the production of knowl- edge. Given that over time the empirical literature has presented evidence more supportive ∗ We would like to thank Norihiko Yamano, Maria Dolores Guillo, Francesco Turino as well as several seminars and congresses participants for helpful comments and discussions. Financial support from Spanish Ministerio of Ciencia y Innovacion and University of Alicante is gratefully acknowledged. All remaining errors are ours. † Corresponding author: Manuela Magalh˜ aes, Department of Economics, University of Alicante, 03690, Spain. Tel: +34 965 903 400 (ext: 3263). Email: mane.magalhaes@gmail.com ‡ Faculty of Economics, University of Porto, and OBEGEF, Rua Dr Roberto Frias, 4200-464, Porto, Portugal. Email:oafonso@fep.up.pt 1