GOAL: To plot a semi-elasticity (percent change in bilateral trade flows with respect to marginal changes in a proxy for "quality of contract institutions") at different levels of another variable (Log of bilateral distance)
PROBLEM: Margins does not calculate the semi-elasticities.
MORE DETAIL:
I am estimating the following extended gravity equation using Poisson Pseudo Maximum Likelihood (PPML) and clustering by country-pairs:
\[ \log{\left(\text{Trade}_{ij}\right)} = \exp{\Bigg\{ \delta_i + \delta_j + \delta_{ij} + \beta_D \Big[\log{\left(\text{Dist}_{ij}\right)}\times\text{RS} \Big] + \beta_C \Big[\text{CE}_i \times\text{RS}\Big] + \beta_{CD} \Big[\text{CE}_i \times \log{\left(\text{Dist}_{ij}\right)} \times\text{RS}\Big] \Bigg\}} \eta \] where i and j index for the exporter and importer, respectively; [\ delta_x \] are fixed effects (exporter, importer and country-pair); RS is an indicator variable for trade in relationship-specific intermediate goods; and CE is a proxy for the "quality" of contract enforcement institutions, which is normalized to take values from 0 (worse) to 1 (best).
PROBLEM: Margins does not calculate the semi-elasticities.
MORE DETAIL:
I am estimating the following extended gravity equation using Poisson Pseudo Maximum Likelihood (PPML) and clustering by country-pairs:
\[ \log{\left(\text{Trade}_{ij}\right)} = \exp{\Bigg\{ \delta_i + \delta_j + \delta_{ij} + \beta_D \Big[\log{\left(\text{Dist}_{ij}\right)}\times\text{RS} \Big] + \beta_C \Big[\text{CE}_i \times\text{RS}\Big] + \beta_{CD} \Big[\text{CE}_i \times \log{\left(\text{Dist}_{ij}\right)} \times\text{RS}\Big] \Bigg\}} \eta \] where i and j index for the exporter and importer, respectively; [\ delta_x \] are fixed effects (exporter, importer and country-pair); RS is an indicator variable for trade in relationship-specific intermediate goods; and CE is a proxy for the "quality" of contract enforcement institutions, which is normalized to take values from 0 (worse) to 1 (best).
