Lecture 7 ECONOMIC GEOGRPAHY AND INTERNATIONAL TRADE By

Lecture 7 ECONOMIC GEOGRPAHY AND INTERNATIONAL TRADE By Carlos Llano, References for the slides: • Fujita, Krugman y Venables: Economía Espacial. Ariel Economía, 2000. • Materiales didácticos de diferentes autores: Baldwin; Allen C. Goodman; Bröcker; J. Sánchez

Index 1. Agglomeration with borders (L is fixed) 1. Border effect: a summary. 2. The Core-Periphery model and international trade 1. An intuitive view. 2. The model. 3. Implications. 3. Application. 4. Extensions. 5. Conclusion

Tema 5 -EE 1. Introduction 1. What is a country? • Some space separated by borders. • What is a border? • Edges behaving as obstacles to movement and information. 2. What is a region? • A space separated by “weak borders” inside the “strong” borders. • There is more factor, goods and service mobility between regions. 3. Has distance died? , Is the transport cost insignificant? 4. What’s the importance of borders as barriers for trade? ; and to migration? 3

Tema 5 -EE 1. Introduction The World is flat Cross Borders in a World where differences still matter 4

Tema 5 -EE 1. Introduction: Border Effect 1. Mc. Callum (AER, 1995): 1. On average, the exports of Canadian provinces to other Canadian provinces are 20 times larger than their exports to an equivalent state in the U. S. (same size and distance). 2. Engels and Rogers (AER, 1996): 1. Evidence from urban price movements suggests that the border imposes barriers to arbitrage comparable to 1. 700 miles of physical space. 3. Gil et al (World Economy, 2005): 1. The Spanish CCAA trade between them is 20 times the trade with other countries. 1. Border effect? , physical differences, accessibility, cultural, legal 5 differences…? , specific effects of sector agglomeration

Tema 5 -EE 1. Introducción: Border Effect 1. How does boundaries influence the mobility of people between countries/regions? 1. Helliwell (AER, 1997): “For every resident in a Canadian province, who was born in a U. S. state, you will meet close to 100 people who were born in some other Canadian province of similar size, distance, and personal income per capita. 2. Conclusion: 1. Boundaries have an important role in the reduction of trade and the movement of people between regions and between countries 2. This importance holds even when the physical boundaries (customs, …) are suppressed (i. e. : Canada 6 USA; borders between European countries) or when

2. The CP Model and International Trade Basic Model: Tema 5 -EE • Until now the Core-periphery Model (FKV, 1999): – 2 regions (north-south); 2 sectors (A agriculture. & Manufacturing) – 1 factor Labor. 2 specializations: agricultural L and manufacturing L. • Only LM is mobile; migration is seeking differences in wages. – There are only transport costs for M: iceberg costs (Trs) • Valid view to explain inter-regional trade and manufacturing agglomeration in certain 7 regions of a country:

2. The CP Model and International Trade Basic Model: Tema 5 -EE • The new Core-Periphery Model (FKV, 1999, Chapter 13): – 2 countries; 2 sectors (A agriculture & M manufacturing) – 2 factors: labor + intermediate goods. • There is no labor mobility between countries ( Lr is fixed). • There is mobility between sectors Lr can move from agriculture to manufacturing, seeking for differences in wages. – There are only transport costs for M: iceberg costs (Trs) • Valid view to explain the manufacturing 8

2. The CP Model and IT: an intuitive view Tema 5 -EE • Without labor mobility there cannot be agglomeration as in the Core-Periphery Model. For it to be possible: – We extend the model by assuming that manufacturing firms produce intermediate and final goods. – We allow L mobility between sectors. • The agglomeration in manufacturing in a country provides “intermediate goods” in abundance and more diversity (Forward Linkages): – Firms that produce final goods that use intermediate goods, buy cheaper and with more variety. • Agglomeration of “final goods” in manufacturing in a country attracts “intermediate goods” manufacturing, (backward linkages): 9 – It can lead to a process of specialization that concentrates manufacturing

2. The CP Model and IT: the model Tema 5 -EE • How does “intermediate goods” work? – Input-Output Model (TIO). – Simply view. Hypothesis: • Manufacturing only uses itself as an input. • The manufacturing good is uses both as an intermediate/final good. • The new production function : – 2 factors: labor + intermediate goods • ls: factor labor in s; • Xrs: input of each variety produced in r the sectors in s CES sum of the intermediate goods inputs 10

2. The CP Model and IT: the model Tema 5 -EE Rather than defining the production function through the “intermediate goods”, we introduce them through the price index: • Price index of the intermediate goods produced in r: – The input composite is a Cobb-Douglas function of labor and intermediate goods. • α (alpha)= share of the intermediate goods. • Price of the input • ns= number of varieties produced in s. ps= FOB price of the variety. Tsr: transport cost. σ(sigma)= elasticity of substitution among varieties of manufacture is the same for firms as it is for consumers The more > agglomeration in manufacturing: + abundance of varieties, < price index of producing 11 intermediate goods and therefore < production costs

2. The CP Model and IT: the model Tema 5 -EE • Sales in manufactures: – Now part of each firm’s output goes to consumers as final consumption (final demand) and part to firms for intermediate usage (intermediate demand). – Er= location r’s expenditure on manufactures Sales in the final demand μ= % that manufactures represent in the expenditure Sales in the intermediate demand • Sales equal the total value of production in region r. • The higher > the number of firms in r, > the larger the intermediate goods demand > the larger the total expenditure on manufactures 12

2. The CP Model and IT: the model Tema 5 -EE • Labor mobility between sectors (A y M). – 2 countries. – Lr= 1. Labor is mobile between sectors. – λr= share of country r’s labor force in manufacturing. – nrprq*r= total value of manufacturing output in country r. – The manufacturing wage bill in country r is a share (1 -α) of the previous quantity: – We choose units such that q*r= 1/ (1 -α) so that:

2. The CP Model and IT: the model Tema 5 -EE • Now we want to direct attention to the allocation of labor among sectors (λr) and to wages: Price index in r=1 Price index in r=2 – Now the price indices in r=1 depend not just on wages in r=1 but also on the price indices of the manufactured goods that are also used as intermediate for other manufactures. 14

2. The CP Model and IT: the model Tema 5 -EE • Wages: Firms make zero profit when the price they charge is such that they sell 1/(1 -α). The wage equations are now: Wages in r=1 Wages in r=2 • Agricultural production depends on agricultural employment (1 -λ). • With this, the income in each country is: Income in r=1 Income in r=2 15

2. The CP Model and IT: the model Tema 5 -EE • • The agricultural wage is the labor marginal product A’(1 -λr). The difference in wages between sectors (vr) will be: Wages in r=1 Wages in r=2 • Labor moves from agriculture to industry if there are differences in wages ( v 1>0) and vice versa. • The long run equilibrium occurs when there are no differences in wages: • Long-run equilibrium manufacturing wages therefore satisfy: • When both sectors operate, wages are equalized. • If there is only manufacturing, it may have w > than the agricultural marginal product. • If there is not manufacturing , w will be less or equal to the agricultural wage. 16

2. The CP Model and IT: el modelo Tema 5 -EE The equilibrium: how does the differences in wages and movements between A↔M influence the variation in the agglomeration of M in a country (λr)? : FOUR FORCES: 2 Stabilizing forces (they induce symmetry; avoid agglomeration) • The response of the marginal product function in agriculture: if the agricultural production function is concave, then ▼LA ▲ the marginal product [A’(1 -λr)] and the Wa. This ▼the incentive for a further movement of labor into manufacturing. • Product market competition: a ▲λ 1 is associated with 17 the supply of more varieties in the country and it reduces the price index G , this shifts the demand curve for each

2. The CP Model and IT: the model Tema 5 -EE There also 2 agglomeration forces (“vertical linkages”: FL + BL) 3. FL (supply): a ▼G 1 induced by the ▲λr reduces the cost of intermediates, tending to increase the instantaneous equilibrium wage. 4. BL (demand): a ▲λr rises the expenditure on manufactures in country 1, shifting firms’ demand curves up and tending to rise the equilibrium: ¡¡¡¡¡IMPORTANT¡¡¡¡¡ 1 er ASSUMPTION: Simplifications formanufacturing wage. • The agricultural production function is lineal in output: A(1 -λ)= (1 -λ) , therefore the agricultural wage =1 and so it is the equilibrium manufacturing wage. • μ<1/2. This is, the level of demand for manufactures is small enough for all of manufacturing to fit in one country, and thus ensures that, even if all manufacturing is concentrated in a single country, this country also has some agriculture. 18 • THERE ARE NO DIFFERENCES IN WAGES BETWEEN COUNTRIES.

2. The Core-Periphery Model: implications High transport costs; T=3 λ 2= “wiggle diagram” manufacturing in country 2 percentage that represents W 1=1 0, 8 λ 1 decreases w 1<1 w 2<1 W 1 Curve λ 2 increases • It shows combinations of (λ 1, λ 2), with wages=1 in country 1. • To the right , w 1<1; and to the left w 1>1 W 2>1 λ 2 decreases Tema 5 -EE W 2=1 λ 1 0 increases 0, 8 λ 1= = percentage that represents manufacturing 19 country in

2. The Core-Periphery Model: implications Low transport costs; T=1, 5 “wiggle diagram” λ 2= = percentage that represents manufacturing in country 2 0, 8 W 2=1 λ 2 Increases Tema 5 -EE 0 W 1=1 λ 1 Increases 0, 8 λ 1= = percentage that represents manufacturing 20 country in

Tema 5 -EE 2. The Core-Periphery Model: implications Intermediate transport costs; λ 2 “wiggle diagram” 0, 8 T=2, 15 W 1=1 λ 2 Increases W 2=1 0 λ 1 Increases 0, 8 21 λ 1

Tema 5 -EE 2. The Core-Periphery Model: implications 1. When is symmetry sustainable (sustain point)? “wiggle diagram” • We assume that manufacturing is concentrated in country 1 (λ 2=0) and (λ 1=2μ). • The concentration of manufacturing in country 1 will be stable as long as this agglomeration is less or equal to the agricultural wage (=1). 2. When does symmetry break and agglomeration is sustainable (break point)? • At the symmetric equilibrium, manufacturing wages in each country equal the agricultural wages (which are = to the marginal product of labor in agriculture). • The equilibrium is stable if increasing manufacturing employment drives manufacturing wages below agricultural, 22 (the DISPERSION economies defeat) causing the movement

Tema 5 -EE 2. The Core-Periphery Model: implications • Solid lines: stable equilibriums; dashed lines: unstable “wiggle diagram” equilibrium. With high transport costs: there is an stable equilibrium (λ 1= λ 2). λ 1, λ 2 λ 1 1 λ 1= λ 2 0, 5 T(B) λ 2 T(S) Making the hypothesis flexible: • The agricultural production function is lineal. • But now μ>1/2. The level of manufactures demand is large. Not all the manufacturing fits in a country if we want wages to be equal in both countries: – 0 1 T If it is attempted to concentrate manufacturing in 1, wages in the country with low manufacturing will rise. When this wage gap arises , some part of the manufactures will remain in 2. • 23 Notice that the curve (λ 2) is

Tema 5 -EE 2. The Core-Periphery Model: implications ω1 , ω2 ω1 1 ω2 1 • • ω1 =ω2 T The concentration in manufacturing in 1, rises the country 1 wages and drives the decline of the ones in 2. Two causes: • Demand in L in 1 increases wages in 1. • Country 1 with more manufacturing varieties have lower prices. Eventually, the wage gap declines with transport costs. 24

Tema 5 -EE 2. The Core-Periphery Model: implications “wiggle diagram” History of the World, Part I (KV, 1995): 1. From an initial position in which the two countries are identical (North and South), an international division of labor spontaneously arises through a process of uneven development. North immediately gains from this division of labor, while South, which suffers deindustrialization, initially loses. The world economy tends to a Core. Periphery structure. 2. Eventually, further reductions in transport cots move the world into a globalization phase: the value of proximity to customer and supplier firms (intermediate L and c) 25 diminishes as transport costs fall, and so the sustainable

Tema 5 -EE 2. The Core-Periphery Model: implications What happens if the importance of the intermediate goods in the production function (▲α) increases ? λ 2= ω1 , ω2 0 ω1 1 ω1 =ω2 ω2 1 • • T This increases the transport costs and it induces a larger wage gap between the two countries. Agglomeration is more persistent and the wage gap is 26

Tema 5 -EE 2. The Core-Periphery Model: implications What happens if the importance of manufactures in consumption in the country (▲μ) increases? ω1 , ω2 ω1 1 ω1 =ω2 ω2 1 • T Now country 2 holds more manufacturing (not all manufacturing fits in country 1). • • The wage gap between the two countries is lower. During the globalization phase, country 1 does not suffer a fall 27 w 1 of (it always grows).

2. The Core-Periphery Model: implications Tema 5 -EE What happens if agriculture has a decreasing returns? Making the hypothesis flexible: • The agricultural production function is decreasing. • μ>1/2. λ 1, λ 2 λ 1 1 λ 1= λ – 2 λ 2 1 An increase of manufacturing labor increases the manufacturing wage but also it increases the agricultural one. T • When T is high (autarky): manufacturing is dispersed. Equilibrium is symmetric and stable. • With intermediate levels of T: there is agglomeration. Symmetric equilibrium is unstable. 28 • When T is low: the symmetric equilibrium is stable again.

Tema 5 -EE 2. The Core-Periphery Model: implications ω1 , ω2 1 ω1 =ω2 ω1 Making the hypothesis flexible: • The agricultural production function is decreasing. • μ>1/2. ω2 1 T 29

Tema 5 -EE 2. The Core-Periphery Model: implications “wiggle diagram” According to these assumptions (when manufacturing is large and agriculture tends to decreasing returns): by reducing transport costs, the equilibrium goes through 3 phases: 1. At high transport costs , the dominant force in determining location is the need to be close to final consumption, preventing any strong geographical concentration of manufacturing. 2. At low transport costs, the dominant determinant of location is wage costs, again mandating dispersed manufacturing to keep labor costs down (it makes that the centripetal forces rule over the centrifugal, the 30 opposite of having intermediate transport costs).

Tema 5 -EE 2. The Core-Periphery Model: implications With no labor mobility “wiggle diagram” but with mobility between countries, between sectors, the concentration in manufactures is possible: 1. Vertical linkages (FL and BL) between industries can lead to a concentration in manufacturing. This concentration can also vary with the transport costs ( and other factors). 2. Not only an uneven distribution of manufactures can rise but also inequalities in wage rates and living standards. 3. Questions: 1. Do these processes have something to do with the division between rich and poor countries while T has reduced? ; 2. Can we expect that a higher European integration leads to a 31 (at least initially) inequality growth?

Tema 5 -EE 2. Application “wiggle diagram” • Diego Puga The rise and fall of regional inequalities. European Economic Review 43 (1999) 303— 334. • ELISENDA PALUZIE, * JORDI PONS² and DANIEL A. TIRADO³: Regional Integration and Specialization Patterns in Spain. Regional Studies, Vol. 35. 4, pp. 285± 296, 2001 32