Economic geography and international inequality Stephen Redding, Anthony J. Venables Department of Economics, LSE Houghton Street London WC2A 2AE, UK Journal of International Economics 62 (2004) 53– 82 Received 25 April 2002; received in revised form 8 July 2003; accepted 14 July 2003

Motivation: Puzzle: The increasing international economic integration the equality in wages, despite of the mobility of manufacturing firms and plants. Potential reasons: endowments,technology, institutional quality, and geographical location. A fully specified model of economic geography (that of Fujita et al., 1999)and cross-country data including per capita income, bilateral trade, and the relative price of manufacturing goods.

Literature review: Model of economic geography: Fujita et al., (1999) Market access and per capita income: Harris(1954), Hummels(1995) and Leamer(1997) Geography and capital income: Gallup et al. (1998) and Radelet and Sachs (1998), Frankel and Romer (1999) Market access and wages for US counties: Hanson(1998)

Stylized facts: Average expenditure on freight and insurance as a proportion of the value of manufacturing imports is 10.3% in US, 15.5% in Argentina, and 17.7% in Brazil (Hummels 1999) The median land-locked country’s shipping costs are more than 50% higher than those of the median coastal country. (Limao and Venables 2001) Access to the coast and open-trade policies ⇒per capita income ↗ over 20% Halving a country’s distance from all of its trade partners ⇒per capita income ↗ around 25% (Gallup et al., 1998; Hall and Jones, 1999; Knack and Keefer, 1997; Acemoglu et al., 2001).

Methodology Develop a theoretical trade and geography model to derive 3 relationships for empirical study. A gravity-like relationship for bilateral trade flows between countries; (market access and supplier access) A zero profit condition for firms; (wage equation) The third relationship is a price index.

Data: cross-section of 101 countries the World Bank’s COMTRADE database and UNIDO industrial Statistics Database

Theoretical framework Wage equation: Fujita et al. (1999) price of country i Gj: price index for manufactures country j Ej : contry j’s total expenditure on manufactures Tij : transport cost factor σ : the own price elasticity of demand

Theoretical framework(2) Bilateral trade flows between countries (trade equation) (supply capacity) (country j market capacity) (Bilateral transport costs between countries) Price Index forms

Empirical framework Trade equation: Market access and supplier access Mi: Supply capacity Si: market capaciy Market access and supplier access

Empirical framework (2) Wage equation for country i wi: wage vi: the price of the internationally mobile factor of production ci: a measure of technology differences, constant A on the right-hand side combines constants from Eq. (8). Price index for manufacturing goods

Trade equation estimation Econometric estimation: Xij: value of exports from country i to partner j ctyi,ptnj: country and partner dummies distij: distance between capital cities bordij: a dummy for whether an exporting country and importing partner share a common border. uij: stochastic error

Construction of market and supplier access: Domestic and foreing part (DMAi and FMAi, resprctively) DMAi(1) and DSAj(1): internal trade costs are equal to the cost of shipping to a foreign country 100 km away and with a common border; DMAi(2) and DSAj(2): link intra-country transport costs to the area of the country DMAi(3) and DSAj(3): the likelihood that internal transport costs are less than international

Wages equation estimation Factor incomes in contry i (related to market and supplier access)

Economic geography and per capita income: Control variables : ( Acemoglu et al., 2001; Gallup et al., 1998; Hall and Jones, 1999; Knack and Keefer, 1997). Hydrocarbons per capita Arable land area per capita Number of minerals Fraction land in geographical tropics Prevalence of malaria Risk of expropriate Socialist rule 1950-1985 External war 1960-1985

Robustness Q: Are the results being driven by the OECD? A: No. (column 7 of Table 3) Q: Are the results being driven by the fact that, even outside the OECD,richer countries tend to be located next to each other? A: No. (column 8 of Table 3) Q: Are our measures of transport costs (distance between countries and the existence of a common border) really imprtant for the results, or is everything being driven by common shocks to GDP across countries? A: Our measures of transport costs are really important for the results.

Supplier access Supplier access and intermediates goods prices Gi: Manufacturing price index of each importing country i SAi: The supplier access of each importing country i

Supplier access Market and supplier access Difficult to seperate the effect of market access from supplier access. Solved by exploiting a theoretical restriction on the relative magnitude of thier estimates coefficients. α: the cost share of intermediates ß: the values of labour’s cost share σ: the elasticity of substitution between varieties

Empirical framework (2) Wage equation for country i wi: wage vi: the price of the internationally mobile factor of production ci: a measure of technology differences, constant A on the right-hand side combines constants from Eq. (8). Price index for manufacturing goods

Economic structure and policy analysis Xij: the value of export from country to partner Yi and Yj: country and partner GDP data llocki and llockj；isli and islj: whether exporting countries and importing partners are land-locked, islands openi and openj: whether exporting countries and importing partners pursue open-trade policies Distij: distance between capital cities Bordij: whether or not an exporting country and importing partner share a common border

Conclusion One of the many potential reasons for the reluctance of firms to move production to low wage locations: remoteness from markets and sources of supply. The effects of economic geography remained highly statistically significant and quantitatively important. The results may seem rather pessimistic for developing countries. ⇒Even if tariff and institutional obstacles to trade and investment are removed, the penalty of distance will continue to hold down the incomes of remote regions. As new markets and centres of manufacturing activity emerge, so the market and supplier access of neighbouring countries improves. Their results point to the importance of understanding the role of geography in shaping the evolution of the cross-country distribution of income.