Complexidade econômica: uma proposta metodológica para identificação de produtos estratégicos

Abstract

This work contributes to the literature on economic complexity literature by proposing a method to identify strategic goods whose production would foster faster economic development. The method, the Multidimensional Matching, introduces two innovations. First, given the central role played by capabilities in the literature, it takes into consideration a large number of socio-economic characteristics prior to considering each good's characteristics. Secondly, it reduces subjectivity in the choice of variables by consistent applying statistical methods. To do so, three statistical models are used sequentially. Principal Component Analysis is used to generate aggregated worldwide indicators on human capital, institutions, government, infrastructure and technology. Next, the predicted probability of export is estimated using a Probit regression to identify a list of strategic goods in the target country without revealed comparative advantages (RCA). Finally, Propensity Score Matching is used to identify countries that have a near predict probability of exporting a given good with RCA while Brazil doesn't. A list of suggested strategic goods is rendered, divided into two hubs. The results suggest that the development of the 19 identified products would make Brazil's complexity rise significantly.