Give credit where credit is due: Tracing value added in global production chains William Powers United States International Trade Commission with Robert Koopman, Zhi Wang, and Shang-Jin Wei February 4, 2011 The views expressed here are solely those of the presenter. This presentation is not meant to represent the views of the USITC or any of its Commissioners.

1 Presentation outline Global value chain: nature and measures Conceptual framework and its contribution –Three important matrices based on block-matrix formulation –Integration of other measures in the literature –Decomposition of gross exports completely into value-added components Empirical results –Highlight regional differences in supply chain participation –Show differences in trade costs from multistage production Database improvements and limitations –Extensions of the GTAP database –Connection to official statistics

2 Value chains, from a product view to a global view What is a global value chain? –A system of value-added sources and destinations within a globally integrated production network Literature –Single product: Dedrick, Kraemer, and Linden (2008) –Single country: Hummels, Ishii, Yi (2001), Koopman et al (2008) –Asian regional chains: Pula and Peltonen (2009); Wang, Powers, and Wei (2009) –Global snapshot: Daudin, Rifflart, and Schweisguth (2010); Johnson and Noguera (2010) –Global time series: Erumban et al. (2010); Wang et al. (2010)

3 Global value chains: Multiple measures Hummels, Ishii, and Yi (2001) measures of vertical trade –VS: share of imported inputs in exports –VS1: share of exports sent indirectly through third countries Newer measures –VAX: domestic value-added in exports (Johnson and Noguera) –VS1*: domestic value-added that returns home (Daudin et al.) aka “reflected” exports Not previously unified in a fully specified framework – turn to this next

4 All output is used as an intermediate or final good at home or abroad Value-added framework: Gross output in a two-country world with N goods, X r : (N×1) Gross output of country r A rs : (N×N) IO Coefficient matrix giving use in country s of intermediates from r Y rs : (N×1) Final demand: Country s’s use of final goods from country r

5 Production system in a two-country world Rearranging, where B sr : (N×N) block Leontief inverse matrix, denoting the amount of total output in s required for a one-unit increase in final demand in country r Y r : (N×1) vector of global use of r’s final goods In block matrix notation

6 Value added in production Value-added shares matrix (2×2N) decomposes value added in production of each sector in all countries Direct domestic value added in production: and where V r : (1×N) domestic value-added coefficient vector; element v ri = 1 – intermediate input share from all countries u : (1×N) vector of ones

7 Value-added exports Exports (2N×2) include both intermediate and final goods (See paper for value-added exports at the product level) Value-added exports matrix (2×2) Fully generalizable to a many-country world

8 Incorporates all value-added measures Vertical specialization: both direct (VS) and indirect (VS1) Domestic value added in exports (VAX) Domestic value added that returns home (VS1*) Indirect (VS1): Country 1’s value added embodied in 2’s and 3’s exports Direct (VS): Foreign value added from 2 and 3 embodied in country 1’s exports Domestic value added in exports (VAX ratio)— includes VS1*

9 Completely decomposes gross exports Further upstream in GVCsFurther downstream Gross exports Domestic value added in exports (VAX) Domestic value added that returns from abroad (VS1*) Foreign value added in exports (VS) Final goods Inter- mediate inputs Indirect exports sent to third countries Exports consumed by direct importer Final goods Inter- mediate inputs Final goods Inter- mediate inputs Indirect value- added exports (VS1)

10 Decomposition of gross exports Advanced economies Other emerging Asia NICs Emerging Asia

11 Supply chain participation: Key differences by region Advanced economies Other emerging Asia NICs Emerging Asia East Asia has the most foreign content in its own exports US uses lots of imported inputs in its exports; imported value supplied by Canada, Mexico, and US itself E. Asia has the longest chains–little of its exported value is absorbed by direct importer (see table 3) Integration in NAFTA makes Mexico an outlier among non-Asian economies Japan sends much of its domestic value to final suppliers indirectly through third countries (see table 3)

Trade costs of multistage production East Asia pays a price for its long chains and relatively high tariffs Advanced economies have low foreign content and, hence, low costs Trade costs on imported inputs Trade costs (tariff + transport), as a share of export value Trade costs on exports 12

13 Database development: Estimating a global Inter-Region IO table Start with 2004 GTAP global trade and prod’n database Add additional detail on source and use of intermediate inputs and final goods (elements of A rs ) Use end-use categories of detailed trade data (HS6) to improve imported intermediate use coefficients –UN Broad Economic Classification (BEC) distinguishes intermediate inputs from final goods in imports from each source in each sector –BEC is better than the alternative: Proportional method assumes the intermediate share in imports from each source is the same as in the home country’s domestic supply

14 BEC shows substantial export heterogeneity Intermediate share of U.S. electronic machinery imports, by source Share from US import use table (54.2%)

15 Benefits and limitations of end-use classifications End-use classifications improve estimates of intermediate inputs entering the importing country Still have to assume proportionality to allocate intermediate inputs to each industry within the importing country –Required data not reported by most national statistical agencies –Problem noted by Committee on Economic Statistics of the American Economic Association (Feenstra et al., 2010) Industry-level estimates of value-added trade may be unreliable with unknown biases, despite their theoretical tractability

Conclusions New value-added framework –Generalizes and harmonizes all measures in the literature –Accounts for the entirety of gross trade –Provides new detail on regional differences in supply chain activity and costs It is now possible to measure trade in value- added terms consistent with official statistics –Ideal database would be consistent with both official trade statistics and national income accounts 16

Questions/Comments? Contact information –Bill Powers –Research Division, Office of Economics –U.S. International Trade Commission –(202)