A multi-country approach to multi-stage production. Jim Markusen, Boulder Tony Venables, LSE

Transcription

1 A multi-country approach to multi-stage production Jim Markusen, Boulder Tony Venables, LSE

2 Extensive evidence on growth of new production patterns in the world economy fragmentation. Questions: What are effects of fragmentation of production on: 1) International production, specialisation and foreign investment? 2) Volume and pattern of trade? 3) Real income, by country and individual? Topic where many possibilities needs synthesis and simplification. Role of applied theory: 1) Richer modelling: new mechanisms 2) Aid to thinking: synthesise possibilities in parsimonious structure 3) Locate regularities equations for estimation - This paper, contribution towards 2 and (perhaps) 3. - Model with minimum number of moving parts that can address issues. - Solution by numerical methods but interpret with standard theory

3 Minimum ingredients for interesting answers? 1) Full general equilibrium who gets the gains? 2) Countries vary in two ways: Factor endowments: Jones et al. Trade costs: FDI, geography. 3) Many countries: At least 3 or a continuum. Get away from the 2-country box 4) Goods and production: At least 3 goods; constant returns. But do not have: Increasing returns/ fixed costs/ firm heterogeneity/ contracting: Antras, Helpman, Melitz.

4 Model: Consumers: homothetic preferences Factors: K and L. Goods: Y and X; C components A assembly Production: constant returns to scale. Factor intensities: By increasing labour intensity; C, X, Y, A. Countries: Continuum of countries distributed on 2 dimensional space: Factor endowments, {L, K}; L [0.1, 0.9], K = 1 L. Trade costs t C = t X = t Y = t A [1, 1.35]. NB: Each country has single trade cost : Same for all goods, all destinations. NB: If import, internal price p Y t ; If export, internal price p Y /t. Compute world equilibrium and illustrate on two dimensional country space tool for synthesising possibilities.

5 Equilibrium production without fragmentation: Autarky above Trade costs specialized in X K abundant diversified specialized in Y L abundant

6 Benchmark: no trade costs (bottom edge only) Figure 1a: No fragmentation: Production shares of GDP Figure 1b: Fragmentation: Production shares of GDP Country i's labor endowment (capital endowment = 1 - labor endowment) QY QX 0.88 Production shares of GDP QC QY QA Country i's labor endowment (capital endowment = 1 - labor endowment) QY QC QAX

7 No trade costs: wage-rental ratio with and without fragmentation 5.00 wage-rental ratio with and without fragmentation wage/rental without frag wage/rental with frag K abundant L abundant WRY WRN

8 0.30 No trade costs: proportional change in welfare due to fragmentation Change in welfare following fragmentation K abundant L abundant Mean change: 0.060

9 Fragmentation: the pattern of production autarky above - - Trade costs CA diversified, C, Y, A YA C CY Y A K abundant L abundant

10 Component production and assembly: Focus on stages of X production Components and assembly Trade costs components K abundant assembly L abundant Location of activities shaped by endowments and trade costs.

11 Component production, assembly and trade: market oriented vs export platform activity depends on endowments and trade costs. Trade costs Components and assembly for home market: (MO) Export components. Self sufficient Import components: Assemble for home market only (MO) Export components: Import assembled goods (EP) K abundant Y only Import components: Assemble for home market and export (EP) L abundant

12 Changes in the volume of trade: Increases for the world, will fall for some countries no change Trade costs increase K abundant fall 2 increase fall 1 increase L abundant Fall 1: Switch from importing X to importing C. Fall 2: Switch from exporting X to exporting C.

13 Changes in real income (dw): Efficiency gains and terms of trade effects. dw = 0 Trade costs dw > 0 dw > 0 dw < 0 K abundant L abundant Real income loss for some previous exporters of X. Within country factor price changes?

14 Analytics: Generically, four possible regimes a country can be in: 1) Autarky 2) Specialisation: only one good produced; good also exported. (3 cases) 3) Partial specialisation: two goods produced, one of them non-traded. 4) Heckscher-Ohlin : two or more goods produced and two of them traded (either import or export). Stolper-Samuelson and Rybczynski effects hold.

15 Heckscher-Ohlin regions: Occurs if two activities both produced and traded in cone of diversification. Regions with Stolper-Samuelson and Rybczynski effects. EG1: set of K/L abundant countries that prod & exp C, and prod & imp Y Countries with high trade costs have higher w/r. Internal price; p C /t ; p Y t; S-S effects (C relatively K intensive) EG2: set of K/L scarce countries that prod & imp C, and prod & exp Y: Countries with high trade costs have lower w/r. Internal price; p Y /t ; p C t; S-S effects (C relatively K intensive).

16 Comparative statics: So far, seen how countries vary in a given world equilibrium: Can also change parameters: Single country: World -- so prices change. EG: Globalization and the growth of world trade 1.80 Average VOT/GDP VOTN VOTY Trade costs relative to central-case values

17 Globalization -- production response to fall in trade costs High trade costs Low trade costs K abundant L abundant K abundant L abundant Point goes from assembly for home market to assembly for export.

18 Summary: - Production regimes depend on factor endowments and trade costs. Can encompass simple theories of horizontal (MO) and vertical (EP) activity Market oriented pattern in countries with high trade costs Export platform pattern in countries with low trade costs/ extreme endowments - Fragmentation reduces trade volumes some regions, raises in others. - Increases volume of world trade as a whole - Real income efficiency and terms of trade: some middle countries lose. Way ahead: - Framework for extending our intuitions on trade theory. Need to be able to do analytics and have intuition outside the H-O cone of diversification. - Single country comparative statics movement between & within regions. - And also need richer modelling: What are the costs of distance? Freight/ time/ face-to-face. What are the barriers to fragmenting? Contracts/ technology transfer.