A Review of Rising Above the Gathering Storm, Part 2

February 28th, 2006

Posted by: Roger Pielke, Jr.

Part 1 of this review focused on Chapter 1 of RAGS. This post focuses on Chapter 2, which is titled, “Why are Science and Technology Critical to America’s Prosperity in the 21st Century?” It seems obvious that science and technology are indeed important to society, and understanding why this is so would be helpful for understanding how to prioritize R&D investments in the context of many other demands on public funds, and the relative desirability different possible R&D portfolios. Unfortunately, this chapter does little more than sandwich reams of information between highly general and simplified assertions of the importance of R&D. RAGS Chapter 2 does very little to answer the question posed in its title. For details, read on.

This chapter begins by simply asserting the answer to the question raised in its title,

The visible products of research, however, are made possible by a large enterprise mostly hidden from public view—fundamental and applied research, an intensively trained workforce, and a national infrastructure that provides risk capital to support the nation’s science and engineering innovation enterprise. All that activity, and its sustaining public support, fuels the steady flow of knowledge and provides the mechanism for converting information into the products and services that create jobs and improve the quality of modern life. Maintaining that vast and complex enterprise during an age of competition and globalization is challenging, but it is essential to the future of the United States.

This series of assertions may seem almost intuitive, and the chapter claims that the relationship of public R&D investments and economic growth are well understood,

“the economic value of investing in science and technology has been thoroughly investigated. Published estimates of return on investment (ROI) for publicly funded R&D range from 20% to 67%.”

However, one of the studies that it cites prominently does not display such confidence or certitude. Scott et al. (2001, available here in PDF) open their report with a telling quote from Georgia Tech’s Barry Bozeman:

In the study of technology transfer, the neophyte and the veteran researcher are easily distinguished. The neophyte is the one who is not confused.

Scott et al. introduce their literature review with a recognition of the challenges faced by scholars trying to understand the complicated relationship of R&D and the economy:

The relationships between public research and innovation are recognised to be an increasingly significant topic in the emerging knowledge economy. However, this is an area beset by high levels of complexity and a surprisingly small amount of empirical research. It is a field where it is easy to be misled by simplistic ideas, or to become confused by such data as do exist and the conflicting interpretations that can be made from them. As this review will show, even now eminent commentators and analysts are grappling with some of the most fundamental dimensions of the relationships between research and innovation, science and technology.

Scott et al. assertion a “small amount of empirical research” does not square with RAGS claim that this area has been “thoroughly investigated.” One might be excused for thinking that RAGS cherrypicked the convenient parts of Scott et al. and ignored the rest. Scott et al. warn the reader that the “intuitive approach” (which RAGS asserts unabashedly) to understanding the role of public R&D in the economy can be misleading:

In the context of limited resources for supporting basic research, and the need to justify the expenditure of these resources, a growing number of policy-makers and academic analysts have become interested in understanding the relationships between basic research and economic activity. Much of this analysis has been underpinned by an attractive intuitive approach to understanding these relationships. This approach is characterised by several logical and sequential steps:

• First, science is mainly seen as a source of new information about how the world works.
• Second, because this information is published openly (as is usual with academic research findings), it is ‘free to all comers’ – a low cost input into economic processes.
• Third, the link between science and technology is obvious: scientific information is used in the creation of new technologies, which are then used in economic activity.
• Finally, given this role of science in the creation of economic returns, it becomes attractive to try to quantify the amount of economic benefit that can be attributed to the basic science elements.

This way of seeing science-technology-economy linkages is so intuitively obvious that for a long time it was simply assumed to be a valid approach. Unfortunately, it contains within it a series of misleading and incomplete ‘mindsets and myths’, the limitations of which have only become apparent through more in-depth investigations in recent years.

Scott et al. are decidedly less sanguine that studies focused on quantifying economic rates of return to research are a useful basis for specific science policy decisions,

Studies that use productivity growth as an indicator of social returns to research investments have a number of problems. In adopting a high level of aggregation in their analysis they rarely control for inter-industry differences in technological opportunity and appropriability. Furthermore, such studies do not reveal how the economic returns are realised and thus do not enable a comparison of the productivity impact of research in different scientific disciplines. A further point to keep in mind is that most measures estimate average rates of return, while marginal rates of return are required for the purposes of resource allocation decisions.

A similar critique can be found in Boskin and Lau (1995). Scott et al. do suggest that R&D provides many benefits to the economy, perhaps even more significantly than narrow studies of economic activity would suggest, through the many “channels” of interconnection between science and the rest of society. They suggest that the management of the relationship of science and society through these channels can be a more useful approach to science policy than by seeking to modulate macro-economic effects in an input-output manner. What is clear from Scott et al. however is that understandings of the relationship of science policy decisions and societal outcomes remain quite murky, unlike the assertions found in RAGS.

RAGS plays fast and loose with the voluminous data that it presents. For instance, RAGS asserts that increasing life expectancy in the United States provides a good indicator of the value of basic research. But this assertion would seem to be countered by the fact that the United States is not even close to first place globally in life expectancy, while countries with longer life expectancy invest far less in health research (and healthcare). The story of life expectancy illustrates the many complexities involved in the relationship of science, technology, and societal outcomes. RAGS presents a large amount of statistical information about how health indicators have improved in the United States over the past century, with the suggestion that these trends were a direct or indirect result of public investments in R&D. This may indeed be the case, but this argument is not developed or made here.

Further, as interesting as it is to see ratios of horses to cars in 1900 versus 1997, it is not clear the relevant of such trivia to the underlying analysis. The most telling conclusion I draw from the various graphs presented about technological progress and market penetration is how spectacularly uncorrelated such trends are with public funding of science and technology. Important questions are raised by thee data, but they are not even touched upon here.

Based on its collection of upward sloping graphs, RAGS takes a page from Bjorn Lomborg’s The Skeptical Environmentalist and Gregg Easterbrook’s The Progress Paradox when it makes the claim that environmental and social indicators are almost universally getting better. It then reiterates its core assumption to explain why we see these improvements:

The science and technology research community and the industries that rely on that research are critical to the quality of life in the United States. Only by continuing investment in advancing technology—through the education of our children, the development of the science and engineering workforce, and the provision of an environment conducive to the transformation of research results into practical applications—can the full innovative capacity of the United States be harnessed and the full promise of a high quality of life realized.

What RAGS has yet to do through Chapter 2 is make an argument in support of this repeated assertion about the importance of R&D. Let me underscore that I also believe that R&D is important, but science policy decision making can and should be based on more than general statements of value. For instance, how might we judge the relative value of one possible R&D portfolio to another? Perhaps RAGS answers this in a subsequent chapter.

One Response to “A Review of Rising Above the Gathering Storm, Part 2”

  1. Markk Says:

    I see, this report has more to do with ‘basic research’ than what I was talking about regarding the first installment. That was more the the D side of R&D. This is even more complicated. What “products” are the fruit of basic ecological or climate research? I think that those points are well made in the review. Even more, if ecological research gave us knowledge to slightly change farming practices, and led to reduced loss of productivity, or more productivity with different crops, how is that factored in? Plus there is the Mongolian Horde card. It doesn’t work most of the time, but it does sometimes work to have a bunch of people to throw at a problem, and you need that baseline number of people that are up to speed in a given area. All these make a model of basic R&D’s relationship to economic productivity pretty difficult. I suspect that this is an area where every model is a “dark theory” as it were, where every time you tested you would get a different result.

    Isn’t that the wrong question anyway? Its seems pretty wrongheaded to make some decision like – “we’ll allocate X dollars to basic research” at all. What “Basic” research? High Energy Physics or infectious disease modeling? They seem pretty far apart to be lumped together. There are only so many hours in a day for people to make decisions on this I guess. At the US national level you got what NIH, NSF, NOAA, NASA, DoD, maybe a half dozen line items, so things really are thrown together.