The point of departure of this thesis was to examine the interplay between the number of exporters and the average exported value per firm by applying the gravity theory on export data from the manufacturing industry from 1999 to 2004. Mayer and Ottaviano (2007) and Bernard et al. (2007) illustrate the difference between domestic and internationalized firms. The firms differ in size, performance and return on inputs. The gravity theory has been highly respected for its ability to explain empirical facts in trade. The gravity equation has been shown to apply equally well for both the new and the old trade theory (Feenstra, Markusen and Rose (1999); Evenett and Keller (2002). It also serves remarkably well when internationalized firms in trade are studied and not only aggregated trade flows.
I have found a significant dominance by the extensive margin from a change in importers’ GDP, distance and importers’ GDP in both the simple and the augmented version. That is, the effect on numbers of exporters from a change in the mentioned RHS variables is larger on the aggregate exports than the effect on average exported value per firm. The extensive margin is also found to dominate the intensive margin when controlling for language, contingent border, EFTA, WTO. The findings strongly support the fact that the number of firms are of importance in international trade. What contradicts the traditional gravity analyzes of extensive and intensive margins in trade are how the effect from EU-membership or being defined as a LDC. This is a somewhat surprising result. Both coefficients have the largest effect on aggregate exports through the intensive margin. But by testing the coefficients with a 5 % level of significance, I find both to be insignificant in M6. An almost similar regression is done by Mayer and Ottaviano (2007) on French and Belgian trade data. They end up concluding that the effect on aggregate exports is mostly driven by effects from the number of exporters.
From the estimation it can be concluded that the number of exports in the manufacturing industry seem to increase in importer’s GDP, presence of Scandinavian language and common border, membership in WTO, EU or EFTA. The firm’s extensive margin is decreasing in importers’ population when the importer is defined as a LDC and in the distance between Norway and the importing country. The average exported value per firm has the same pattern, but is reported to be decreasing if the importer is a member of EU, WTO or EFTA. However, the dominance of extensive margin in aggregated exports is statistically significant only through the GDP, population and distance variables.
As Bernard et al. (2007) claim, my findings support the newer theories of trade. The large influence on aggregate exports by number of exporters supports the hypothesis of heterogeneous firms in trade. When distance increases, only the most efficient firms find it profitable to export since the distance implies higher transport costs. On the other hand, if importers’ GDP increases, new exporters enter the market. The new trade theory assumes love of variety which implies that adjustments in export happen on the intensive margin if distance increases. I find no support for this hypothesis, except when controlling for EU and LDC. But then again, their coefficients are found statistically insignificant.
The literature regarding internationalized firms is still young. Further investigation is necessary but also exciting. The information about how the firms’ intensive and extensive margins impact aggregate exports draws a more detailed picture of the relationship between determinants in trade. Further investigations of manufacturing industry could be to decompose the aggregated exports in the manufacturing industry in a more sophisticated way. For example, it could be possible to divide into the number of exporting firms, number of exporting product per firm and average value per exported product per firm. The decomposing of number of products could be defined from the NACE system (a European industry standard classification system). Bernard et al. (2007) and Mayer and Ottaviano (2007) have done this on trade data from European countries. They argue with strong evidence that the extensive margins – both number of exporters and number of exported products – play a significant role when the aggregated level of exports are determined.