Greg Tassey's new paper pulls together and augments a number of themes and recommendations that he has espoused and championed over the past decade. During that period, he has been an unabashed critic of the "vast majority of economists in the United States" who espouse a "neoclassical economic philosophy" that is "inaccurate," because of dependence on outdated views of comparative advantage based upon shifts in relative prices. Of most concern to Tassey, who is chief economist of the Commerce Department's National Institute of Standards and Technology, is an alleged indifference to allowing industrial technology "to offshore in response to trends in comparative advantage" and the lack of policies to maintain "a viable domestic-based manufacturing capability." Indeed, though he alludes to the necessity for more broad-based economic growth models, Tassey does not provide such a comprehensive alternative: rather, the bulk of the paper is concerned with one (quite important) component of economic growth: technology policy, as a subset of broader R&D and innovation policy. He argues for a new public–private paradigm in this area that will create policies to manage the "entire technology life cycle," including all major elements of domestic and global supply chains.
Implicit in his critique of current "neoclassical" economics is a deep-seated skepticism of the effects of globalization, particularly the impact of outward foreign direct investment (FDI) by US multinational companies, and--most important--the movement to locate R&D facilities outside the United States. The paper underscores repeatedly the purported danger from the fact that R&D investment by US multinationals outside the country is growing at three times the rate of domestic R&D investment. "This trend needs to be reversed," he argues.
In the brief space allotted for this comment on the paper, I will not be able to pass judgment on all of the facts and arguments advanced in the dense and detailed analysis. Indeed, one preliminary note on presentation would be that the paper could have been cut by at least half in order to sharpen Tassey's most important points. I found a slide presentation of the paper a good deal more cogently argued. In any case, whatever my questions and points of disagreement, I would urge all those interested in US R&D policy and the future US manufacturing to read and absorb Tassey's analysis and conclusions.
That said, reviewers are expected to have questions and reservations; and here are mine. First, the indictment of so-called neoclassical economists is over broad and to some degree a caricature. Specifically, regarding R&D policy that is the intense focus of this paper, for some decades all economists (neoclassical or Keynesian, to list a branch of economics Tassey excoriates) have come to accept a much more complex explanation of the innovation process. No economist really believes today (if ever in the past) that "new technologies will magically appear from advances in basic science." Indeed, in the early 1990s, Richard Nelson and Nathan Rosenberg, two of our most distinguished economists of science, described a more complex view of innovation when they noted (as opposed an older, linear model associated with Vannevar Bush): "To say that new technologies have given rise to new science is at least as true as the other way around. And it is more on the mark to say that with the rise of modern science-based technologies, much of science and technology has become intertwined."
Though there is no fully developed new paradigm of the innovation process even to this day, and various names have been attached to alternate theories, so-called "chain-linked models"--not unlike the descriptions of the process advanced by Greg Tassey--are widely accepted. In this formulation, the chain of innovation consists of a series of diverse actions and responses that produce multiple feedbacks and links between the underlying core of scientific knowledge and invention, testing and production--and even the need for new scientific breakthroughs before completing the process for a given product or group of
products. And Nelson, in an analysis from the same time period, belies Tassey argument that economists did not and do not "recognize the complex and integrated character of modern technology systems" In an essay revisiting Bush Report after 50 years, he described the complex dynamics of the innovation, observing that "in virtually every technology and every industry there in fact has evolved over the years a division of labor between companies in the industry, and universities and other public institutions, and between private and public funding of the relevant research that feeds into the advance of technology. . . . And what makes sense in one industry may not in another."
And therein lies the nub of one great difficulty in the model Tassey advances: he decries current growth strategies because of their "inability to understand the complexity of the typical industrial technology and the synergies among tiers in high-tech supply chains." But this very complexity and the huge diversity of the public and private division of labor among industries noted above by Nelson creates a high skepticism regarding the confident assertion that the US must emulate other nations and "create comparative advantage through technology." On the public side, which technologies--either proprietary or generic--should be chosen? Tassey argues for a "portfolio of investments," based upon "systematic planning that identifies new technology trajectories early and then implementing institutions to manage them in the same way a mutual fund is managed." (He calls for a new National Innovation Foundation). In addition, Tassey posits that "sound management" includes "the expectation that individual R&D projects will fail and be terminated." In sum, the new model "destroys the convention wisdom. . . that government funding of technology research is 'corporate welfare' or 'picking winners or losers." But as a number of studies have shown--including the oft-cited, classic analysis by Roger Noll and Linda Cohen--once there is a bureaucratic commitment (even for "proof of concept" or demonstration) backed by private resources, project termination is problematic. In the end this new paradigm ends up closely resembling an old model: a government entity, staffed by a bureaucracy immune from the realities of the market will have to make bets on future technology payoffs dependent on private partners with a stake in the game and a desire to augment their investment with public resources.
On a broad technical level, Nobel Laureate Robert Solow summed up the policy
conundrum flowing from imperfect competition and scale returns leading potentially to first-mover advantage and long-term dominance in key technologies--a central goal for Tassey--when he remarked (paraphrasing):
I know that there are lots of industries where there are four dollars worth of social output to one dollar of private input; my problem is that I do not know which one they are.
And on a practical level, this led economist Jagdish Bhagwati to add that "it is very hard for policymakers, and very easy for lobbyists, to decide which industries have the externalities."
Finally, a word about the pervasive fears and warnings about the current and future impact of the outward FDI and R&D outsourcing by US-based multinationals. Many economists do not share Tassey's gloomy assessment of FDI and the impact of "offshoring." For instance, recent work by Mathew Slaughter of Dartmouth College and earlier a member of the President's Council of Economic Advisers, directly counters the claims that US multinationals have "abandoned" the US and that their global expansion "tends to 'hollow out'" their US operations and "export jobs abroad."
Among the points of rebuttal Slaughter advances are the following. First, the operations and investment of US multinationals are still highly concentrated in the US itself: 71.6% of output; and 74.3% of worldwide capital investment.
Second, of particular importance regarding the fears expressed in the Tassey, US
multinationals perform just under 87 per of their R&D in the US--or 6.5 dollars of
knowledge investment for every 1 dollar spent abroad. While foreign-based R&D will undoubtedly increase over time, interviews with CEO of US technology companies demonstrate that they are well aware of the dangers of placing their high-end "crown jewel" research activities overseas--and seem to be successfully resisting pressure even from large technologically rapacious countries such as China to trade technology for market access.
Third, affiliate sales are overwhelmingly in the host country or other foreign markets. In 2006, the latest year for data, Sales by US-owned affiliates were over $4.1 trillion, of which just $280 billion was imported back into the US (roughly 93 vs. 7%). To a great degree, they do not come back to the US and displace US workers.
Fourth, foreign affiliates are still located largely in high-income countries, not lowerincome economies (about 79% in 2006, with China and India, recent overall fast GDP growth, only 1.8 and 0.5%, respectively).
One could go on through other evidence by Slaughter and others (US wages for workers in multinational firms are higher than the rest of the private sector, and the net income of foreign affiliates now rivals that of US parents, demonstrating a close link between growth and profitability abroad and at home). As to "neoclassical" blindness to the consequences of global competitive imperatives, unfortunately for Tassey his new bosses in the Executive Office of the President--certainly no bastion of market fundamentalism--seem cut from the same misguided mold as the "majority of (US) economists." The recently released Economic Report of the President, written by the president's Council of Economic Advisers and headed by Christina Romer, an economist who has written extensively about the US economy, cites ideas of Adam Smith and David Ricardo with approval and notes particularly that "specialization" is a "primary source of trade-related productivity" and that "Ricardian comparative advantage. . .can be seen in a number of aspects of US trade."
The report also describes the major changes in international trade through "vertical specialization" where firms have production in multiple countries. By 2001, 75% of US manufacturing trade was intra-industry trade either with outside firms or with foreign based affiliates. Though the new supply chains can complicate analysis of comparative advantage, the Report concludes that: "Specialization by process should allow the United States to focus on jobs oriented toward the processes that match the human capital, physical capital, and technology in the United States (thus) increasing productivity." As with other neoclassical treatments, the Report also acknowledges that not all elements of
society will benefit equally from specialization and that adjustment policies must be put in place to buttress the social safety net.
While this reviewer has reservations about major elements of Greg Tassey's new
technology policy model, there are areas of substantial agreement. I fully agree with the call to legislate a permanent R&D tax credit at a fixed level, rather than the incremental form that the US has adopted in the past. "Temporary" tax credits, whatever their field or goal, are antithetical to rationale corporate planning and thus in this case will inevitably frustrate the intended goal of inducing great R&D investment. I also agree with his call for increased--and steady state--federal R&D spending. Thus, I would applaud, as does Tassey, President Obama's commitment to raise federal R&D spending to a uniform level
of 3% of GDP. Third, though I would redefine the term, there is certainly a case for
targeting "long-term breakthrough research." For the perspective adopted here, that would mean support for what Richard Nelson and others have called "targeted basic research"--what Tassey in the paper refers to as Pasteur's Quadrant--though the "demonstration projects" he lists need further clarification and I am wary of the suggestion that public funds (as venture capital) should go to companies to overcome the "risk spike." Fourth, there is a solid public good component to government programs that foster the spread of information about science and technology advances and best practices, including a manufacturing extension service, particularly for small and medium-sized firms.
In the end, whatever the differences on policy, I applaud the depth of research and analysis Tassey has brought to this effort. It will certainly advance the debate on the complex roles of many actors, both public and private, in the evolving national and global innovation process.
Claude Barfield is a resident scholar at AEI.