50th Congress of the European Regional Science Association: "Sustainable Regional Growth and Development in the Creative Knowledge Economy", 19-23 August 2010, Jönköping, Sweden
The distance puzzle has been wildly discussed in the literature since Leamer and Levinsohn (1995) shed the light on it. This puzzle simply says that "the world is not getting smaller": distance still matters to account for trade. This is reflected in a decreasing distance of trade (DOT), or in a stable or increasing (negative) elasticity of trade with respect to distance (Disdier and head, 2008). Several explanations of this puzzle have been emphasized in the empirical International Economics literature. Duranton and Storper (2008) contribute to this issue in offering a full theoretical framework to account for this puzzle. They explain that an improvement in transport technology can increase the transport costs, because it creates an incentive to trade higher quality goods. Our paper proposes a new theoretical explanation of this puzzle. This explanation shares with Duranton and Storper's two important characteristics: i) there is a non monotonic relationship between transport cost and trade cost. ii) this phenomenon is due to contract incompleteness. However, the mechanism that we underline is quite different: in our model, based on a Dixit-Stiglitz increasing return to scale technology, a fall in transport cost increases the international division of labour. It follows that input-output linkages require a higher level of coordination. Such a coordination is easier between neighbors than between very distant countries. As a result, trade increases with all partners, but more quickly for neighbors than for distant countries.