EUROPEAN UNION AND TRADE INTEGRATION: DOES THE HOME BIAS PUZZLE MATTER?

This paper analyses the size of the border effect or home bias within the European Union (EU) with the aim to quantify its impact in the trade integration process which started in 1992. The gravity model serves as a general framework where the use of sectoral data allows obtaining a more accurate measure of the border effects. The econometric analysis applied to a gravity model that contains a high disaggregation in the data introduces heteroskedasticity problems together with the presence of zero values, therefore, in order to obtain consistent estimates, the use of the Poisson Pseudo-Maximum Likelihood estimation method is recommended. Our results show that the bilateral trade flows depend upon the size and proximity of the trade partners, together with other variables as adjacency or language. Evidence in favour of a positive and diminishing border effect has been found for the period 1995-2006 and it accounts for 20% to 22% within the EU-19.