ABSTRACT
Tables 2.5 and 2.6 present the results of estimating Equation (2.1) separately for energy and non-energy trade. One would expect the gravity considerations to be quite a bit weaker for energy trade than for non-energy trade. To some extent this is the case. Focusing on the pooled results of Tables 2.5 and 2.6, one can see that the R2 is considerably lower for the energy trade of Table 2.5 than that of the non-energy and total trade in Tables 2.6 and 2.4, respectively. The effects of the contiguity and same nation dummy variables, though still positive, are no longer statistically significant. The biggest single change is in the coefficient of the FTA dummy variable, which is positive and significant for non-energy trade and otherwise for energy trade (Tables 2.5 and 2.6). Yet, in other respects the results are surprisingly similar. Indeed, the influences of the same colonizer dummy, distance, and exchange rate volatility variables are even stronger for energy trade than non-energy trade. The latter finding may be due to the fact that companies involved in shipping, refining, and marketing of oil are owned by investors or governments of the former colonial powers and that transport (other than by pipeline) of oil and especially natural gas is particularly costly.
