Iterative solutions for structural gravity models in panels

Aurélien Poissonnier

International Economics, 2019, issue 157, 55-67

Abstract: The estimation of structural gravity models for bilateral trade flows faces methodological difficulties. These difficulties are in particular due to the multilateral resistance terms (MRT) which account for the general equilibrium constraints of global trade. In the present paper, I show that the MRT are unique (to a harmless scale factor) and one can compute them iteratively. Based on a parallel between the MRT and the biproportional projection of matrices in the Input-Output literature, I give new insights on the econometric strategy to estimate structural gravity equations. I extend to panels the iterative method proposed by Head and Mayer (2014), Gravity equations: Workhorse, toolkit, and cookbook, in Handbook of International Economics. This double loop procedure nests the current standard in the literature (the Poisson Pseudo Maximum Likelihood –PPML-with dummies). I exemplify its possibilities on real data. PPML results appear not robust to a change of estimator.

Keywords: Structural gravity model; Panel data; Multilateral resistance; Biproportional projection method (search for similar items in EconPapers)
JEL-codes: C13 F14 (search for similar items in EconPapers)
Date: 2019
References: Add references at CitEc
Citations: View citations in EconPapers (6)

Downloads: (external link)
https://www.sciencedirect.com/science/article/pii/S2110701717300434 (text/html)

Related works:
Journal Article: Iterative solutions for structural gravity models in panels (2019)
Working Paper: Solving for Structural Gravity in Panels Yes We Can (2016)
This item may be available elsewhere in EconPapers: Search for items with the same title.

Export reference: BibTeX RIS (EndNote, ProCite, RefMan) HTML/Text

Persistent link: https://EconPapers.repec.org/RePEc:cii:cepiie:2019-q1-157-4

Access Statistics for this article

More articles in International Economics from CEPII research center Contact information at EDIRC.
Bibliographic data for series maintained by ().