A major obstacle in the estimation of inter-industry (vertical) productivity spillovers is the need of an exogenous productivity shock. In this article, I propose the use of trade policy changes as a large productivity shock, since the literature has found evidence that trade liberalisations increase industry-level productivity. To do so, I develop a new empirical methodology using spatial econometrics, and apply it to the large economy-wide shock represented by the 1989–1998 Brazilian trade liberalisation. My results indicate the existence of positive and substantial upstream productivity spillovers. Nevertheless, no evidence of downstream productivity spillovers is found.
ASJC Scopus subject areas