Alejandra Martinez

Working Papers

Trade relationships during and after a crisis: Evidence from road disruptions in Colombian flower exports
I study the impact of an extreme weather event on international trade relationships. Exploiting variation in Colombian flower exporters' access to cargo terminals during an unprecedented La Niña season in 2010-11 when some roads became impassable, I find that exporter-importer relationships exposed to road disruptions became 7 percentage points less likely to end during the road disruptions. These results are driven by importers that can hedge against non-deliveries by relying on current relationships. A firm-level exposure measure shows that relationships linked to importers who cannot rely on other current relationships for sourcing are more likely to end. I present a theoretical framework that rationalises the decision to keep or replace a relationship based on (1) the relative cost of establishing a new trade relationship; and (2) firm profits affected by other exposed relationships. The findings shed light on the dynamics of international buyer-seller relationships in the context of extreme weather events.

Transitivity in trade
(With Carlo Perroni and Dennis Novy.) A large literature on social networks has shown that individuals are more likely to be connected with each other if they share connections with other individuals. We take this idea to trade between firms. Firms are more likely to trade with each other if they share common trading partners, and forming a new connection makes trade with other partners more likely. We provide a statistically based characterisation of transitivity in trade transactions that is applicable to any model of trade where the formation of agent-to-agent links involves a stochastic component (e.g., arising from random search or random matching), and we discuss how transitivity can be measured in transactions data. We apply this methodology to firm-level Colombia-US exporter-importer data.

Work in Progress

Insights from firm-to-firm transaction data: Reassessing buyer-seller networks