State Dependence and Unobserved Heterogeneity in the Extensive Margin of Trade

From MaRDI portal
Publication:100784

DOI10.48550/ARXIV.2004.12655arXiv2004.12655MaRDI QIDQ100784FDOQ100784


Authors: Julian Hinz, Amrei Stammann, Joschka Wanner Edit this on Wikidata


Publication date: 27 April 2020

Abstract: We study the role and drivers of persistence in the extensive margin of bilateral trade. Motivated by a stylized heterogeneous firms model of international trade with market entry costs, we consider dynamic three-way fixed effects binary choice models and study the corresponding incidental parameter problem. The standard maximum likelihood estimator is consistent under asymptotics where all panel dimensions grow at a constant rate, but it has an asymptotic bias in its limiting distribution, invalidating inference even in situations where the bias appears to be small. Thus, we propose two different bias-corrected estimators. Monte Carlo simulations confirm their desirable statistical properties. We apply these estimators in a reassessment of the most commonly studied determinants of the extensive margin of trade. Both true state dependence and unobserved heterogeneity contribute considerably to trade persistence and taking this persistence into account matters significantly in identifying the effects of trade policies on the extensive margin.








Cited In (1)





This page was built for publication: State Dependence and Unobserved Heterogeneity in the Extensive Margin of Trade

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q100784)