ECO364 - International Trade The Organization of Production Christian Dippel University of Toronto Summer 2009 Christian Dippel (University of Toronto) ECO364 - International Trade Summer 2009 1 / 21
Introduction Introduction The fragmentation of production processes has increased dramatically in the last 15 years. Computerization has made coordination easier and has led firms to increasingly outsource some of the steps in the production process to specialized suppliers of intermediates. In addition, MNE increasingly source these intermediates from suppliers in other countries through offshoring. At the same time the operation of foreign affiliates through Foreign Direct Investment (FDI) has also increased dramatically. Christian Dippel (University of Toronto) ECO364 - International Trade Summer 2009 2 / 21
Introduction Some Facts Trade in services and intermediates has increased much faster in the last 10 years than trade in final products. In addition, FDI of both horizontal and vertical type have increased much faster than trade in final products. In bilateral trade flows, a larger share of imports is intra-firm if exporters are more capital-abundant. At the sector-level, a larger share of imports is intra-firm in sectors that are more capital-intensive. Christian Dippel (University of Toronto) ECO364 - International Trade Summer 2009 3 / 21
Introduction The Multinational Firm (MNE) These patterns suggest that the fragmentation of production is intimately linked to a firm s organizational choices In the last chapter, we saw horizontal FDI as a location choice based on a trade-off of closeness to market and the additional fixed costs of setting up multiple plants There was no organizational choice - firms were always integrated. This is different when we consider vertical FDI. Christian Dippel (University of Toronto) ECO364 - International Trade Summer 2009 4 / 21
Introduction A Firm s Decision Problem The empirical patterns suggest a two-dimensional decision problem: Whether to integrate or outsource part of the production process Where to locate production Christian Dippel (University of Toronto) ECO364 - International Trade Summer 2009 5 / 21
Introduction Christian Dippel (University of Toronto) ECO364 - International Trade Summer 2009 6 / 21
Theory of the Firm Theory of the Firm The Decision of Integration vs. Outsourcing has been explored in the IO literature on the theory of the firm The following outlines the so-called property rights view of the firm: A good z(j) is assembled from headquarter services h(j) and an intermediate component m(j): z(j) = h(j) η m(j) 1 η To simplify the discussion of trade, we assume that z(j) is itself an intermediate and that the final good x(j) can be produced one-to-one from z(j). We assume h(j) and m(j) cannot cross borders and only z(j) is traded and the final good is produced in the location of the headquarter. The final good producer supplies h(j) and decides whether to produce m(j) within the boundaries of his firm or outsource to a specialized supplier (who is assumed to have a productivity advantage) Christian Dippel (University of Toronto) ECO364 - International Trade Summer 2009 7 / 21
Theory of the Firm Incomplete Contracting The 1st key assumption is that both inputs are completely specific to good j and are therefore worthless outside of this relationship. The 2nd key assumption is that there is incomplete contracting which means that the two suppliers engage in ex-post (after production) bargaining over how the revenue should be split up. This creates a bilateral hold-up problem. The match creates total revenue R(j) = p(j)h(j) η m(j) 1 η This is a two stage game in which both sides make their production decisions in the first stage, knowing there will be bargaining in the second stage Christian Dippel (University of Toronto) ECO364 - International Trade Summer 2009 8 / 21
Theory of the Firm The Hold-Up Problem We assume Nash-Bargaining in which each party receives their outside option plus shares β and 1 β of the R(j) Because each side receives only a share of total revenue, they both under-invest in their respective inputs (relative to the joint revenue-maximizing levels) Christian Dippel (University of Toronto) ECO364 - International Trade Summer 2009 9 / 21
Theory of the Firm Organizational Choice The key distinction between integration and outsourcing is in the outside option: With integration, there is still a hold-up problem but the final-good producer is able to appropriate a share δ of the revenue if bargaining breaks down. The outside option is therefore δr(j) and the final good producer receives a share β V = δ + (1 δ)β > β of the revenue under Nash Bargaining. Christian Dippel (University of Toronto) ECO364 - International Trade Summer 2009 10 / 21
Theory of the Firm Optimal Bargaining Weights Why would a final good producer outsource if he receives a higher bargaining weight under integration? Because a higher bargaining weight makes increases the supplier s incentive to under-investment (even if this supplier is integrated in the firm) - the final good producer gets a larger slice of a smaller pie. There is an optimal bargaining weight from the final good producer s perspective This β trades off the under-investment incentives of both sides β depends critically on η - the more important the headquarter services are in the production of the final good, the higher is the optimal bargaining weight β Christian Dippel (University of Toronto) ECO364 - International Trade Summer 2009 11 / 21
Theory of the Firm The Nash bargaining weight β reflects power-relations/contractual laws and can not be affected by the final good producer But the final good producer can choose between integration and outsourcing This choice depends on whether β or βv are closer to β The optimal choice of organizational form depends on η β is a monotonically increasing non-linear function of η At low levels of headquarter-intensity (low η), outsourcing is optimal while at high levels, integration is optimal. Christian Dippel (University of Toronto) ECO364 - International Trade Summer 2009 12 / 21
Theory of the Firm Christian Dippel (University of Toronto) ECO364 - International Trade Summer 2009 13 / 21
Trade Trade Let s embed this model in a many country-2sector-2factor Heckscher-Ohlin World Let there be two sectors: H (high-tech) and M (manufactures) H is headquarter-intensive and M is component-intensive: η H > η M Suppose headquarter-services h(j) are capital-intensive and components m(j) are labor intensive This makes the high-tech sector capital-intensive and the manufacturing sector labor-intensive. Recall that only z(j) is tradable. What sorting patterns does the theory predict? Christian Dippel (University of Toronto) ECO364 - International Trade Summer 2009 14 / 21
Trade High-tech firms will integrate and manufacturing firms will outsource because ηh lies above the cutoff where integration is optimal and η M lies below that cutoff High-tech plants will locate in capital-abundant countries Manufacturing plants will locate in labor-abundant countries Within-firm imports will therefore be higher from capital-abundant exporters countries. Within-firm import will be higher in capital-intensive sectors. Christian Dippel (University of Toronto) ECO364 - International Trade Summer 2009 15 / 21
Trade Christian Dippel (University of Toronto) ECO364 - International Trade Summer 2009 16 / 21
Trade and Wages revisited We concluded our trade-and wages chapter by noting that SBTC looks like a far more likely candidate than trade for explaining increased wage-inequality in OECD countries. This was commonly agreed upon by economists in the late 90 s. We noted that Directed Technological Change might play a large role in OECD countries but can not explain increasing wage-inequality in developing countries because the TC should be unskill-biased there. Let us now consider trade-in-intermediates as an alternative explanation, based on Feenstra and Hanson (1999) Christian Dippel (University of Toronto) ECO364 - International Trade Summer 2009 17 / 21
Trade and Wages revisited Let us assume a HO world in which endowments of North and South are outside the Cone of Diversification so that FPE fails and countries are specialized (not diversified). Suppose consumers derive utility from consuming a unique final good Y which is assembled from a continuum of intermediate inputs, indexed by i [0, 1]: Intermediate inputs x i are produced from high- and low-skill labor We maintain the HO assumption of identical technology (identical a H (i) and a L (i)) but allow for a difference in Total Factor Productivity (TFP) of North and South φ N > φ S. Christian Dippel (University of Toronto) ECO364 - International Trade Summer 2009 18 / 21
Trade and Wages revisited If FPE fails, then wages are not equalized and skilled labor will be relatively cheaper in the country that is more abundant in skill: wh N/w L N < w H S/w L S In such a non-diversified world, North and South will each produce the range of intermediate goods in which they have a cost-advantage This is similar to our analysis of the multi-good Ricardian model. Christian Dippel (University of Toronto) ECO364 - International Trade Summer 2009 19 / 21
Trade and Wages revisited Suppose we rank goods in increasing order of a H (i)/a L (i) (skilled-labor-intensity). Then the relative cost of producing good i in the North rather than in the South is decreasing in i there is a unique cutoff i such that the North produces all goods with i > i and the South all other goods Christian Dippel (University of Toronto) ECO364 - International Trade Summer 2009 20 / 21
Trade and Wages revisited Now consider an increase in φ S so that the South becomes more efficient (this can reflect reduced trading frictions) The effect is to raise i The North loses it s least skill-intensive intermediates to the South where they are the most skill-intensive. The key is that - because goods are ranked in skill-labor-intensity - this increases the relative demand for skilled labor both in the South and the North Key idea is that international outsourcing is associated with a de-location of production processes that are unskilled-intensive from the point of view of the North, but that are skill intensive from the point of view of the South. This raises w S /w U in both North and South. Christian Dippel (University of Toronto) ECO364 - International Trade Summer 2009 21 / 21