Innovate to Export or Export to Innovate? Evidence from Chile

TL;DR: In a 2014 article in Estudios Públicos with Eduardo Bitran, Cristián González, and Marcelo Villena, we test the direction of causality between innovation and exporting in Chilean manufacturing firms over 1995–2010. In most subsectors, export propensity Granger-causes greater innovation effort — participation in global markets pushes firms to innovate, not the other way around. Minerals/base metals and metal-mechanics show mutual causality; wood and paper show none.

Why does the direction of causality matter?

There is broad consensus that exporting firms are more productive than non-exporters (Bernard & Jensen, 1999), that innovation raises firm productivity (Crépon, Duguet & Mairesse, 1998), and that exporting itself can stimulate R&D (López Rodríguez & García Rodríguez, 2005). Exporting, innovation, and productivity are tightly linked — but which causes which? If innovation comes first, policy should subsidize innovation and let exports follow. If exporting comes first — the learning-by-exporting channel — then export promotion is the more efficient instrument. The literature also documents mutual reinforcement, where both directions operate at once (Filipescu et al., 2013), so the question is empirical and must be answered case by case.

What does the paper do?

The paper builds an econometric model for Chilean manufacturing subsectors over 1995–2010, controlling for each firm's productivity level. Two design choices distinguish it from earlier studies: it uses total innovation expenditure rather than only R&D, capturing a wider range of productivity-seeking activities inside the firm, and that broader measure allows a larger sample of firms to be studied.

What did we find?

Using Granger-causality tests by subsector:

• In most subsectors, export propensity causes greater innovation effort — global market participation comes first, innovation follows.
• Minerals and base metals, and metal-mechanics show mutual Granger causality: exporting and innovation reinforce each other.
• Wood and paper show no statistically significant causality in either direction.

What does this mean for policy?

For an economy like Chile's, the results favor instruments that push firms into global markets — the discipline and learning that come with exporting trigger innovation spending. A policy mix that treats innovation subsidies as the sole lever would miss the stronger channel running from exports to innovation.

The full text, citation, and BibTeX are on the publication page. The article appeared in Estudios Públicos, issue 134 (Fall 2014), pages 109–130.