Quantifying the Impact of Exporter-Specific, Importer-Specific and only Time-Varying Variables in Structural Gravity

Advances in gravity literature have presented econometric approaches for the theoretically consistent estimation of structural gravity. When estimating the impact of policy-shocks on trade values however, researchers are confronted with two problems. Once multilateral resistances are taken into account, through time-varying importer and exporter fixed effects, they absorb the effect of policy-shock indicator variables. Hence, we cannot obtain a coefficient for the impact of policy. The second problem is rooted in the necessary panel data dimensions in structural gravity that requires multiple-exporters and multiple-importers. The (at least) three dimensional panel implies that any coefficients/impacts that are estimated apply to the whole set of exporters rather than the country related to the scope of the research. I propose a method to approach these two problems, estimate the impact that policy-shock variables have on trade and differentiate the results for the country/countries related to the scope of the research. A short application on the impact that the Global Financial Crisis had on trade values is presented.