The Chicken or the Egg: Causality Between Trade and Innovation

While the literature is quite clear on the association between trade and innovation, there has been little explicit study of the direction of causality. This study uses all patents granted in the USA between 1987 and 1999, assigns them to probable industries of origin and sectors of use, then tests causality with trade flows in those same economic sectors. We run robustness checks on various measures of trade (imports versus exports, volume versus value of trade), and on various measures of innovation (patent counts, patent claims, citation-weighted patents, and patent counts weighted for originality or generality). Results at the aggregate level support the literature’s assumption that imports cause innovation which then leads to exports, but at the industry level, causality is more complicated, running in both directions from imports (or exports) to and from innovation.