Discussion of Francisco Moris U S International Trade

Discussion of Francisco Moris’ U. S. International Trade in R&DRelated Services and a Transactions-Based Profile of Business R&D’ Wolfgang Keller, University of Colorado and NBER

Overview • Shows recent trends in US trade in research, development, and testing (RDT) services • Develops a new methodology, transactions-based R&D accounting, for tracking international R&D that emphasizes ‘use’ of R&D results • Applies the new framework to 2003 US data from NSF and BEA

Why is this important? • Because the US trade surplus in RDT services is primarily due to exports by foreign-owned multinational affiliates ? • New indicators on how new technology is used and diffused internationally for current levels of integration • R&D creates new technology, thereby raising productivity – and that is the critical determinant of per-capita income

Sources of Income Differences q. Production function: Y = A*F(K, H, L) Two ways of raising output: 1. Higher effort and investment – More hours worked: ΔL – More investment: ΔK, ΔH 2. Higher efficiency – More output for given L, K, and H: ΔA

Output Differences: the Breakdown • Typical example: U. S. output per worker in the year 1985 is 35 times as large as that of Niger Output per Worker difference Capital difference 1. 5 x Human capital difference Productivity difference 3. 1 x Total 35 Source: Hall and Jones 1999 7. 7 =

But: not much is known yet about the ‘A’ factor • Typically, what we know is – Not based on data, but on constructed or otherwise estimated data – Only correlations of observables with the constructed ‘A’ factor • Some indications that ‘A’ is affected by technology activity both at home and abroad • There is a great need to measure and quantify technology creation and linkages

Research, Development, and Testing Services

US: trade surplus in Business, Professional , and Technical Services (BPT) and RDT Services

US is net exporter of RDT largely through affiliated trade… Millions of current $ US Total Unaffiliated Affiliated

And that comes largely from foreign-owned affiliates to their multinational parents

US RDT trade surplus through exports by foreign -owned affiliates –does this matter? • Quantitatively, how large are these net technology transfers from the US to other countries? – Not that large: 1. 5% of US BERD in 2003 • Are there policy implications for the US? • Technology sourcing, ‘Listening-post FDI’ – van Reenen et al. , forthcoming American Economic Review • International technology linkages go both ways. FDI coming into the US has a major positive effect on US firm productivity – Keller and Yeaple (2005)

Who is the ultimate beneficiary of US R&D? • Current R&D expenditure statistics focus on two dimensions (Frascati Manual) – R&D Performer – R&D Funder • System of National Accounts - based statistics would consider the R&D ‘user’ • If market transactions on R&D services are observed, we know their value, not only the costs of R&D

Relation to new US R&D Satellite Accounts Frascati SNA Manual (FM) R&D Output expenditur es Exports and imports of -Exports of R&D RDT output -Imports Capital of RDT formation See Robbins (2006)

The three faces of R&D Funding Production Use/Exchange

Table 4: A New Transactions View of R&D Transactions 2 to 5 are market-based

R&D in the closed economy Gross Expend. R&D used Note: [2] = [4], [3] = [5]

Trade in R&D services in open economy • R&D service exports of one country are the R&D service imports of another: [2] + [3] = [4’] + [5’] (assumes no intracountry R&D trade)

Application to 2003 US Business R&D [1]+[6]+[4]+[5] [1]+[6] +[2]+[3] Units: Billion current $ US

Initial questions: domestic R&D trade • In which industries is there a lot of R&D service trade? • What are the characteristics of firms that buy and sell R&D services domestically? • How large are R&D purchases and sales, relative to own-account R&D in the closed economy?

International R&D service trade • What are the characteristics of firms that export or import R&D services? With which other countries? • Does R&D service trade decline with distance, domestically as well as internationally, like most other trade does?

Technology diffusion: market-based and nonmarket based

Why look at Multinationals: FDI and the Internalization Question • Internalization : The hierarchical structure between MNE parent and affiliate overcomes market failures – internalize externalities--in the market for technological knowledge -> that is a motive for FDI, versus technology sale or licensing

Studying technology diffusion inside and outside the MNE • Comparing RDT service trade between – affiliated and – unaffiliated parties • Are there systematic differences in these patterns? • If so, does transfer pricing play a role? • To the extent that it does not, what accounts for the systematic differences?

Conclusion • The paper provides the methodological basis for new research on technology linkages between countries – Will improve understanding of international technology diffusion & income differences • Current efforts should also put emphasis on royalty and licensing payments • And: how can a more informed analysis of market-based linkages shed new light also on spillover relations?