Estimation of the Gravity Equation in the Presence of Zero Flows

Product Heterogeneity, Intangible Barriers and Distance Decay: The Effect of Multiple Dimensions of Distance on Trade across Different Product Categories

“…6 Robustness Linders and de Groot (2006) show that sample selection (in their case zero flows) of aggregate trade flows might have a limited impact on the coefficients. To test the robustness of our classification on more disaggregated trade flows regarding sample selection bias, we therefore repeated the finite mixture procedure using only 'important trade flows'.…”

“…Heckman, 1979): namely, zero flows of the dependent variable, here bilateral trade, may not be random. Nevertheless, using aggregate trade data, Linders and de Groot (2006) show that OLS does not greatly suffer from selection bias. We therefore use a straightforward regression framework.…”

The Canola Oil Industry and EU Trade Integration: A Gravity Model Approach

“…The flow analyzed here is the import of canola oil for non-food use (TARIC: 15141110) into EU countries (EU Export Helpdesk, 2009 However, since this zero-inflation can be treated as a selection bias problem, it can be resolved using the method of Heckman (1979) as advised by Linders & de Groot (2006). Among the possible specifications, Martin & Pham (2008) prompt to use the 2-step-Heckman approach for this specific case.…”

Determinants of the Korean agricultural trade with the LDCs and the OECD countries

“…However, problems with the traditional models owing to statistical issues, such as omitted variables and dropped observations, have led the more recent studies to develop the applied gravity research. Estrella (2012) and Linders and de Groot (2006) pointed out that in the case of zero values of trade, it is not possible to use the logarithmic form and that a heteroskedasticity problem arises. To overcome these problems, they used the HTS analysis.…”