The issue of convergence of per capita income has been debated extensively by economists since the seminal works of Baumol [9], Barro & Sala-i-Martin [8] and Mankiw et al. [36] on examining the predictions of the Solow model ([51]). These studies were based on the Solow-Swan growth model. Solow [51] and Swan [53] introduced the convergence hypothesis as part of the neoclassical growth models. These models proposed that the initial per capita income is poor countries must show negative linkage to growth in per capita income for both rich as well as the poor countries. Nevertheless, there have been few studies on trade convergence in the EU, an interesting issue in the theory of economic integration and macroeconomic harmonization.
In the relative economics literature, there are ongoing debates as to whether trade integration accompanies highly correlated business cycles. As concerns the linkages between trade and business cycle correlation, Krugman [34] and Eichengreen [22] have provided evidence of a negative association between more intense ties and business cycle synchronization. Kalemli-Ozcan et al. [31]-[32] found analogous result and they argued that due to economic integration, the increased opportunities for income diversification may lead to higher specialized production and thus higher business cycle asymmetry. Kose & Yi [33], using a basic business-cycle model have also suggested the existence of a negative linkage between larger trade flows and cross-country symmetry of macroeconomic variability, while Imbs [27] reported results indicating there is not a significant association between increased trade and business-cycle synchronization because trade integration affects national economies through several channels. However, some other authors have provided different results by focusing on international spillover effects. More specifically, in the seminal study by Frankel & Rose [23], the development of international trade was found to have a strong positive link with synchronization of cyclical fluctuations. In addition, De Haan et al. [20], Bordo & Hebling [16], Inklaar et al. [28], Zervoyianni et al. [57], Anastasiou [3], Zavou et al. [55] and Calderon et al. [17] and [18] reached analogous conclusions.
Furthermore, Ben-David [9] and Sachs et al. [44] stressed that open economies experienced unconditional convergence. Sarkar [46] using panel data for 51 countries for the period 1981–2002, reported results indicating a positive linkage between openness and growth under the assumption of highly trade dependent countries. The Sarkar [46] finding was confirmed by the study of Billmeier & Nannicini [13] after controlling for endogeneity. Bernhofen [12] concluded that there was a strong relationship between convergence of economics and trade development. However, to the best of our knowledge, only one study ([29]) has tried to provide evidence on the association between the role of trade (both intra-EU and EU versus the world) and government expenditures with the per capita income convergence in the EU during the period 1995–2017. Their results implied that there is a strong positive link between trade openness and government expenditures with per capita income convergence in the EU.
None of these studies, to the best of our knowledge, has concentrated on the convergence in the trade volumes in European countries. Only Radimersky & Hajko [41] employed beta convergence in the export volumes for only SITC 6 and 7 trade categories and they concluded that the speed of unconditional convergence of the export volume per capita were about 0.05–0.06.
This paper explores for the first time how we can improve the convergence in exports and imports volumes in European environment (EU28 member states, 4 EU candidate countries and Switzerland) by considering a three-stage procedure. Initially, we employ absolute beta convergence for both trade volumes and then one solution to improve the convergence, could be to intelligently identify the countries that negatively contribute to the convergence, eliminate them from the equation and rerun the procedure. More specifically, the countries that are the most different from the others could be probably eliminated, so the convergence not only will be prevented but will be further improved. Then, in order to achieve the above, we consider the Euclidean distance approach. Thus, the purpose of this paper is twofold. First, it attempts to add to the existing literature by examining the exports and imports annual data for the presence of convergence in the European environment for the period 2010–2019. Second, we propose a new three-stage convergence procedure by considering beta convergence and Euclidean approach in order to improve the trade convergence in Europe.
The rest of the paper is organized as follows. Section 2 describes the literature on association between trade, growth and business cycle synchronization, In Section 3, we proceed to describe the new procedure applied. Section 4 reports and discusses the results of the econometric approaches and Section 5 contains concluding comments.