45th Congress of the European Regional Science Association: "Land Use and Water Management in a Sustainable Network Society", 23-27 August 2005, Amsterdam, The Netherlands
Does trade lead to increased cross-country regional business cycle synchronization and reduced national economic borders? The theory does not really provide an unambiguous answer. Our paper addresses empirically this question using Argentina and Brazil as case studies of developing countries. These countries liberalized unilaterally trade since the mid-1980s and also established MERCOSUR (a regional integration agreement with Paraguay and Uruguay) in 1991. As a consequence, the intensity of trade between Argentina and Brazil rose significantly. The answer to the initial question is no. The increase in bilateral trade between Argentina and Brazil did not translate into significantly more synchronized regional business cycles. Using Gross Provincial Product for Argentina and Gross State Product for Brazil for the period 1961 to 2000, we find that within-country regional business cycle synchronization is substantially larger than cross-country regional business cycle synchronization. Moreover, this difference has increased over time. These results are mainly driven by Argentina's behavior and hold even after controlling for factors such as distance, size, sectoral specialization, and the degree of regional fiscal policy coordination. The empirical evidence based on Brazilian states and Argentina as a whole suggests that the higher level of trade among regions within a country is an important factor to that accounts for the observed border effect. In the case of Argentina additional factors such as monetary and exchange rate policies and large country-specific shocks have also played a significant role.