Abstract
The paper applies the gravity model of international trade in its analysis of German exports. The added value of our research is derived from the innovative shift in focus from the traditional gravity model specifications to the national level in order to interpret its estimations in a non-traditional way, but remain consistent with data structure and thus bring new insights into the analysis of German export performance.
Our panel dataset includes German exports to 176 countries and 22 control variables including institutional factors over the period 1995-2011. We estimated a Random Effects model and also a Least Trimmed Squares model to control for the heterogeneity between countries.
We distinguish two panel data specifications: time-series and cross-section. This allows us to examine long-term and short-term decision horizons.
The general conclusion of our model is that German exporters are more prone to expand the trade to countries that are more distant from their European neighbourhood relative to the world average. Exports are sensitive to both the real exchange rate movements and the price levels of partner countries, even though their elasticity is significantly less than unity, which suggests that German exports would not be impacted very much if the Euro appreciated in real terms.
The position of the Euro in German trade seems to be rather ambiguous since not all tests revealed its role as a catalyst.