Do imports of intermediate inputs generate higher productivity? : evidence from Ecuadorian manufacturing firms

International trade has been stated as one of the most important mean of improving firms' productivity, being the channel behind, the technology transfer from foreign companies to local firms. Focusing on imports, they can positively contribute to local firm's productivity performance by incorporating better inputs in their production processes. This paper analyzes, by using an augmented Cobb Douglas production function, the effect of imported intermediates on the production level of manufacturing formal firms in Ecuador, and, the causal relationship between the import decision and firm productivity. For this, we make use of a unique administrative data ranging from years 2007 to 2018 and estimate the total factor productivity (TFP) at firm level using a flexible Levinsohn-Petrin semiparametric method, which allows us to address the endogeneity and simultaneity problem existing in the inputs selection (inherent to this type of estimations). The results show that (i) a 100% decrease in the share of domestic intermediates in total intermediates increases productivity by 7% in the manufacturing sector, (ii) when we use a measure of import intensity, we find that a 100% increase in the share of imported inputs increases firm productivity by 7%, (iii) foreign and domestic intermediates are substitute inputs, with an elasticity of substitution positive and greater than one (4.43), and (iv) we provide robust evidence in favor of the "learning-by-importing" hypothesis. Finally, we find that there is self- selection of more productive firms into the import market. The main conclusion is that the decision to import foreign intermediates improves productivity in manufacturing firms. This matters from a policy perspective as increasing productivity levels is crucial for a country's economic growth.