| Abstract [eng] |
The main purpose of this master thesis is to investigate the effects of economic size and trade costs on international trade over different periods, based on the gravity model. The work consists of three main parts: the analysis of literature, the research and its results, conclusion and recommendations. The literature review examines the construction of the gravity model, its evolution, and the work of other researchers. It was revealed that the first intuitive gravity model was based on Newton’s theory of gravity, which states that bilateral trade is positively influenced by the GDP of the exporter and importer and negatively affected by the distance between them. Researchers later sought to adapt this model to trade theory, eventually developing a universal structural gravity model suitable for empirical research. This model incorporates other factors affecting trade costs, such as shared borders, regional agreements, and others, instead of merely the distance between countries. Scholars use the model to study the trade potential of individual countries as well as bilateral trade between specific groups of countries, such as the OECD or the EU. Additionally, the gravity model can be applied not only to the overall analysis of countries’ exports and imports but also to the examination of international trade in specific sectors or product groups. It is also an effective tool for analyzing the impact of short-term shocks on foreign trade. During the research, intuitive, structural, fixed-effects, and Poisson pseudo-gravity models were constructed for both cross-sectional and panel data. The analysis revealed that statistically significant results can be obtained using both intuitive and structural gravity models, and no multicollinearity issues exist. However, it cannot be concluded that the obtained estimates are efficient, as all models exhibit heteroskedasticity. Therefore, based on literature recommendations, the analysis transitioned from the ordinary least squares (OLS) method to the Poisson pseudo-maximum likelihood (PPML) method. After constructing the Poisson pseudo fixed-effects gravity model, the previously obtained estimates changed slightly but did not alter the conclusions: the distance between countries, whether the countries are neighbours, whether they share a common language, colonial ties, or have had colonial relationships, and the existence of trade agreements all influence bilateral trade between countries. Nevertheless, the Poisson pseudo gravity model provides contrasting insights compared to the fixed-effects model evaluated using OLS. According to the PPML model, the impact of colonial ties on trade becomes increasingly significant over time, whereas the influence of regional agreements on international trade diminishes. The conclusions and recommendations synthesize the findings from the literature review and empirical analysis, providing guidance for future research and policy measures. |
