Introduction

Trade or transport costs are a key element of new economic geography (NEG) models in determining the spatial distribution of economic activity (see e.g. Krugman 1991; Venables 1996 and Puga 1999). Without trade costs there is no role for geography in NEG models. It is therefore not surprising that trade costs are also an important ingredient of empirical studies in NEG (see Redding and Venables 2004; Hanson 2005 or Head and Mayer 2004). They are a vital ingredient of a region's or country's (real) market potential, which measures the ease of access to other markets (Redding and Venables 2004; Head and Mayer 2006). In the empirical trade literature at large, trade costs are also a main determinant of the volume of trade between countries (see e.g. Limao and Venables 2001; Anderson and van Wincoop 2004).

The empirical specification of trade costs is, however, far from straightforward. Problems with the measurement of trade costs arise because between any pair of countries they are very hard to quantify. Trade costs most likely consist of various sub-components that potentially interact, overlap, and/or supplement each other. Obvious candidates are transport costs, tariffs, and non-tariff barriers (NTBs), but also less tangible costs arising from cross-border trade, due to institutional and language differences for example, have been incorporated in previous studies (Limao and Venables 2001). An additional difficulty arises with what is arguably the most obvious measure of trade costs, transport costs.