1
Introduction
In recent decades, production processes of intermediate and final products have been increasingly fragmented across countries in what are called global value chains (GVCs). GVCs may involve companies in one country outsourcing stages of production to unrelated entities in other countries, multinational enterprises (MNEs) offshoring stages of production to units of the MNE overseas, or both. GVCs can also involve completely independent companies merely sourcing their parts from whichever upstream company may be the most competitive, with no control arrangement necessarily involved. The changing global trade environment and the changes in firms’ behavior have raised new and more complicated issues for policy makers and have made it difficult for them to understand the extent and operations of GVCs and their spillover effects on national and local economies.
As providers of information to policy makers, researchers, and the public, national statistical organizations need to respond to the increasing prevalence of GVCs. However, while they publish a great deal of information on economic performance, trade, employment, and foreign direct investment (FDI), they have found it challenging to collect or present these data in a way that quantifies the international linkages and flows that characterize GVCs. National accounts are designed to measure economic activity within national borders. Asymmetries in accounting standards and data collection among countries further exacerbate the problems of gaining an accurate picture of economic activity and production, as well as market flows, within and across borders.
The increasing prominence of GVCs requires improved statistical frameworks, including those for business and trade statistics, to better reflect the increasingly complex economic activity that characterizes how goods and services are produced. Of particular importance is trade in research and development (R&D) services and other intangibles for which measurement is less straightforward and coordination across countries is necessary to produce comparable cross-country statistics. While the data for understanding trade in goods are robust, less is understood about trade in services, enterprise
characteristics, and intangible assets. In the past, measurement was less complicated—production and R&D were conducted within the same firm, and a statistical agency could send the firm a survey. Now large firms produce and innovate in different enterprises and locations.
To improve the understanding, measurement, and valuation of GVCs, the National Center for Science and Engineering Statistics (NCSES) asked the Innovation Policy Forum at the National Academies of Sciences, Engineering, and Medicine to organize a workshop. A planning committee requested, reviewed, and selected papers to present at the workshop, which was titled “Innovation, Global Value Chains, and Globalization Measurement” and was held May 5–7, 2021. The research presented at the workshop built on a number of reports and conferences—involving the efforts of NCSES, the Organisation for Economic Co-operation and Development (OECD), and statistical organizations in many other countries over the past 15 years—on the intersection of globalization and intangibles and the challenges for official statistics. The project’s full statement of task is in Box 1-1.
ROLE OF THE PLANNING COMMITTEE
This proceedings has been prepared by the workshop rapporteurs as a factual summary of what occurred at the workshop. The planning committee’s role was limited to planning and convening the workshop. The views contained in the proceedings are those of individual workshop participants and do not necessarily represent the views of all workshop participants, the planning committee, or the National Academies of Sciences, Engineering, and Medicine.
OVERVIEW OF THE WORKSHOP
The workshop presentations summarized in this volume cover a variety of data sources from a variety of countries, as well as a number of policy questions. The goal for the presentations was to inform future research and data development and to help inform national statistical organizations and international collaborations working to improve measurement of R&D globalization. The workshop presentations focused principally on a series of papers invited by the planning committee and were presented by one or more of the authors of those papers, who in some cases also drew on their related work.1Appendix C describes how the papers contribute to better understanding GVCs, their measurement, and related policy issues. Additionally, the keynote address focused on challenges and prospects in GVC measurement methodology, and the panels discussed lessons learned from the workshop.
___________________
1 References to presentations in this volume give the name of the presenting author(s). At times, they were joined in the session by additional coauthor(s), some of whom participated in question-and-answer portions of the session.
Among the topics addressed in the presentations by invited paper authors were the relationship between production and innovation (Chapters 2, 5, and 13); the impact on trade of the treatment of capital through FDI (Chapter 3) or as an intermediate input (Chapter 12); and whether countries’ tax policies affect intellectual property and trade measurement (Chapter 4). Two of the paper presentations dove deeply into a particular industry—wind energy (Chapter 6) and automobile components (Chapter 7).
A number of presentations focused on the issue of colocation of production and innovation. Gumpert (Chapter 2) described how she and her colleagues used data from Bureau van Dijk’s Orbis database combined with information on intellectual property from the European Patent Office’s (EPO’s) PATSTAT database to investigate an MNE’s choice of locations for production and innovation. They found that, compared with basic innovation, applied innovation is more likely to be colocated with production because of the synergies between applied innovation and production. They also found that basic innovation is more likely to be offshored to more advanced economies where human capital resources are likely to be more abundant, and that applied innovation is more likely to be offshored to emerging markets where production is cheapest.
Similarly, Fan (Chapter 5) described his use of the Orbis database and EPO’s PATSTAT database to explore offshoring of R&D by MNEs, finding that the innovative intensity of an affiliate increases when the affiliate is located in a country with a higher-quality talent pool. He discussed his observation that innovation and production tend to be colocated.
Fort’s presentation (Chapter 13) also explored colocation of production and innovation, based on her coauthored paper. Her presentation described the coauthors’ look at U.S.-based companies using Census Bureau data to examine the relationship between manufacturing and innovation, specifically to determine whether colocating production near innovation increases innovation. While she and her coauthors found that colocating innovation near production does increase innovation, former manufacturing firms that no longer produce products continue to innovate. In fact, the share of patents from nonmanufacturers has grown over time and is almost half of all patenting activity.
Wei (Chapter 3) described his and his colleagues’ use of OECD Analytical Activities of Multinational Enterprises (AMNE) data to account for the contribution of FDI in estimates of GVC activity. He showed estimates that do not account for FDI miss around half of what could be considered GVC activity, when accounting for value-added generated through FDI.
Similar to Wei, Ding’s presentation (Chapter 12) described his use of tables from the World Input-Output Database (WIOD) to investigate whether treatment of capital inputs as an intermediate input affects trade statistics. He found that gains from trade are much larger when capital is considered an intermediate input.
Oddo’s presentation (Chapter 4) described his and his colleagues’ use of Italian tax data to determine whether there is mismeasurement of trade due to companies’ shifting intellectual property products (IPPs) to tax havens. While they found strong evidence of such mismeasurement, their discussant cautioned that it is difficult to move IPPs and that such movements are usually overstated.
Two paper presentations dove deeply into particular industries. Surana and Doblinger (Chapter 6) described their and their colleagues’ look at GVCs for wind technology using worldwide data from Navigant to find information on component manufacturing of wind technologies. Although previous researchers have found that outsourcing innovation has a deleterious effect on long-run innovation, Surana, Doblinger, and colleagues did not find such an effect; however, they did find a difference in geographic location of innovation depending on whether the firm was a component supplier or an original equipment manufacturer (OEM). They also found that component suppliers are likely to patent short- and long-term innovations regardless of whether they are in advanced economies or emerging markets, but that OEMs in advanced economies are more likely to hold long-term patents than OEMs in emerging markets.
In her presentation, Helper (Chapter 7) described her and her coauthor’s use of U.S. customs data to explore firm-to-firm relationships in GVCs pertaining to the manufacture of components of automobiles. They explored whether firms organize their supply chains in response to organizational strategy. In such cases,
firms may enter into purchase agreements not only to reduce transactions costs, but also to create tight-knit supply networks to build credibility and take advantage of economies of scope.
Tybout (Chapter 11) cautioned about the use of U.S. customs data. His presentation revealed that there may be issues in the accuracy of these data, although such microdata are essential to understand firm-to-firm transactions.
Warzynski (Chapter 10) described his and his coauthor’s use of Danish data to investigate whether a firm will purchase from a firm in the same or another country. They used data collected by Statistics Denmark and surveys of offshoring coordinated with Eurostat. They found that while there has been an increase in outsourcing to the global South, this outsourcing has become more diffuse over time, with Danish firms outsourcing to firms in many more countries. They also found that MNEs are more likely to outsource to their affiliates.
Bircan’s presentation (Chapter 9) described his and his colleagues’ combination of geocoding from patent data from EPO, the U.S. Patent and Trademark Office, and the Japan Patent Office with Orbis data to explore the diffusion of innovation from MNEs. They found that barriers to diffusion may be related to time zone and physical differences. Gender also apparently matters—female investors appear to have less mobility than their male counterparts.
The keynote presentation at the workshop, by Van Reenen, focused on extensive spillovers to domestic firms from MNEs (Chapter 8), especially in high-R&D industries. According to Van Reenen, this work suggests the existence of a sales certification effect—that domestic suppliers that supply superstar MNEs tend to see increases in their sales, even outside of their sales to the MNE.
The penultimate session of the workshop (Chapter 14) focused on challenges for measuring GVC broadly, for modernizing official business statistics, and for measuring R&D and competitiveness for U.S. exports. The discussant (Francisco Moris) placed the research presented at the workshop within the context of extensive international efforts to develop and update guidance and standards for business and trade statistics, as well as measurement of R&D and intangibles. He proposed an integrated framework for identifying high-, medium- and low-priority projects to research, develop, and implement improved statistics on R&D, trade in services, GVCs, and relevant aggregates for the System of National Accounts.
In the final workshop session (Chapter 15), a panel was asked to identify lessons learned from the workshop. Individual panel members singled out the policy needs for statistics to understand the effects of GVCs on trade, different industries, and within-country economic growth or decline; the importance of finding innovative ways to share business data for national statistics and research, including linkage of microdata; and the use case for more detailed information on firm location and ownership, particularly for MNEs.
REMAINDER OF THIS VOLUME
The remainder of this proceedings volume is divided into chapters summarizing the keynote, paper presentations, and final panels in the order in which they occurred during the workshop. Chapters 2 through 13 summarize the presentations of workshop papers and the keynote. Chapter 14 summarizes a panel discussion focused exclusively on measurement issues, and Chapter 15 summarizes a final panel with reflections on the lessons derived from the workshop. These chapters are followed by a list of references that appear in the volume and three appendixes: Appendix A reproduces the workshop agenda, and Appendix B includes biographies of workshop speakers and members of the planning committee. Appendix C describes how the papers contribute to better understanding GVCs, their measurement, and related policy issues.
