Our guest blogger, Christian Ross, is a PhD student in the School of Life Sciences at Arizona State University and a National Science Foundation Graduate Research Fellow. His research is focused on the intersections of science, society, and science policy, particularly surrounding emerging biotechnologies.
There is a recurring problem that I see in ongoing discussions about public funding of scientific research. And, perhaps surprisingly, this problem is not specific to the current US administration1, but rather seems to be endemic in scientific communities as much as political ones. Often, science policy is reduced to a simple question: do we need more basic research or more applied research? I think that this distinction between basic and applied research is unnecessary and unhelpful, particularly for biotechnology and biomedicine. Insistence on the basic-applied distinction as the question in science policy obscures growing evidence of the ineffectiveness of all publicly funded science to deliver on its promises to the public to provide life-improving technologies and products.
Thus far, the Trump administration has not made many friends within scientific communities. Threatened and realized budget cuts to research funding across disciplines and what is often called “anti-science” rhetoric have led to many impassioned responses from scientists and science advocates proselytizing about the importance and necessity of public funding of scientific research, especially basic research. But what counts as basic research, and more importantly how useful it actually is, nearly always goes unexamined.
Back to Basics: The Linear Model
The prevalence of the basic and applied research distinction in funding conversations in the US has strong roots in the work of Vannevar Bush and his report to President Harry Truman on the role of science in post-WWII America, Science: The Endless Frontier. Basic research, as opposed to applied research, focuses on pursuing fundamental understanding of natural phenomena without direct consideration about the potential applications of that knowledge to create novel technologies. Applied research, by contrast, harnesses the scientific knowledge generated by basic research to create technological solutions to societal problems in engineering, medicine, and other applied fields.
While the distinction between basic and applied research by no means began with Bush and Science: The Endless Frontier, Bush’s work did serve to legitimize the basic-applied distinction as a foundational aspect of post-war US science policy. Bush argued that national military and commercial interests depended on good science. And good science was basic science, unhindered by the constraints of applications. Moreover, basic research was the “pacemaker for technological progress” and invariably led to applied technologies and societal benefits.
Bush’s work also solidified an implicit social contract between science and society. In return for public funding, science would provide technological solutions and products to improve society through advances in healthcare, sanitation, travel, communication, national defense, or economic outcomes. Since then, federal funding for science has been characterized as supporting either basic or applied research with consistent pressure from scientists to keep basic research as the essential fuel for scientific and societal advancement.
Deceptively Linear Beginnings
Bush’s “linear model” for scientific development has already been widely and frequently critiqued (including at times by this blog) for oversimplifying the technical and political complexities of scientific research and development. Those are fair criticisms, and I am not intending to pile on much more here. Actually, I hope to defend the linear model a bit (or at least those that have bought into its ideology) as an understandable misstep. Though, I only want to do so to show that even if a linear model approach to science was once helpful or fitting2, it is not so for our present moment nor going forward, especially not in the fields of biotechnology and biomedicine.
That said, it is understandable how the linear model gained credibility in the US, particularly in biotechnology and biomedicine. Leading up to and during WWII, US applied research focused heavily on developing basic physics research (especially atomic physics) to support the war effort. After the war, there was a surplus of basic research in virology, cell biology, and genetics that had not yet been applied to developing technologies and products. This swell of untapped basic research directly led to the development of many new biotechnologies and drugs with the advent of recombinant DNA technology in the 1960s3. Recombinant DNA techniques enabled scientists to modify bacterial genomes to produce complex biochemical compounds for use in new, synthetic pharmaceuticals, like synthetic insulin to treat type I diabetes, erythropoietin (EPO) which treats chronic kidney disease, and tissue plasminogen activator (tPA) which breaks down blood clots to treat strokes. Much like the linear model describes, basic research acted as a precursor to biotechnological and biomedical applications.
The development of these new drugs through bioengineering was an unqualified success and provided justification for continued use of a linear model approach to science funding.
Biotech Paradise Lost?
However, we do not see similarly large leaps forward in biomedicine and biotechnology today. To be sure, we regularly develop new therapies, drugs, and techniques, but not at rates (or profit margins) comparable to the 1970s and 1980s4. And we certainly are not making progress on the high-priority problems like cancer in the same way as early biomedical and biotechnology research did with renal disease and diabetes.
According to linear model thinking, a slowdown of technological development is the result of a lack of basic scientific research. But over the past fifteen years federal support of basic research has been at least double what it was in the 1970s through the mid-1980s. Evidently, increased funding for basic research has not translated into increased biomedical and biotechnological development. Even if we grant that the linear model at one time adequately described the relationship between federal funding of science and basic and applied research5, it is clearly not doing so now.
So, what changed? Why does the linear model no longer seem to describe the developments of biotechnological and biomedical innovation? Put simply, the reasons and context that enabled scientific and industrial boom in biomedicine and biotechnology in the 1970s and 1980s are different now in three major ways.
First, there was an unusual backlog of untapped basic life science research available for application by industry. During WWII and early years of the Cold War, basic research in virology, cell biology, and genetics seldom translated into applied research. During that time, the biological sciences did not receive nearly the same attention as other fields, like physics, resulting in much of its basic research to remain unconnected to tangible applications. However, once recombinant DNA techniques were developed and their potential became apparent, applied research in biotechnology and biomedicine grew dramatically. Researchers took advantage of surplus of basic biological research to jumpstart the applied research of the biotechnology and biomedicine industries in the US. New basic research became incorporated into the applied research and products pipeline as quickly as it could be published.
Second, new biotechnology and biomedicine industries and markets emerged. Before the 1960s, the life sciences and industry were more separated from each other. There was some overlap in medicine, but nothing of the magnitude that came with the developments in the 70s and 80s. For the first time, biotechnology and biomedicine startup companies emerged that proved highly profitable in a newly created market. Today, academic and corporate researchers now more readily recognize biotechnology and biomedical research as sources of technical problem solving and profit.Corporations are already established and dominate what once was a more open, competitive marketplace. There simply are not the same commercial opportunities for startups to be innovative in the lab insulated from the pressures of industry like there used to be.
Third, the problems that could be solved with biotechnology and biomedicine in the 1970s and 1980s were relatively simple and straightforward. The advances in the technical capabilities of biotechnology and biomedicine made accessible an entirely new class of biological problems that previously had been beyond the scope of science. In this newfound class of biological problems there were problems ranging from relatively easy to devilishly difficult. Understandably, researchers at new startup companies first picked the low-hanging fruit6.
Now, the biotechnological and biomedical problems that exist are technically and socially more complex than those of the mid-twentieth century. Although the development of recombinant drugs was certainly challenging, the biotechnological and biomedical problems that exist today by comparison are much more difficult. Also, because the problems are harder, research ventures to develop commercially viable products are more costly and more risky for both academic and corporate researchers.
Further, we are more aware that we work within larger, more complex systems with greater ranges and degrees of uncertainty. Science at every stage is hard, and it makes sense to solve the easiest problems first. But that means that once the relatively easy problems have been solved, only harder, often wicked problems remain.
Now What Do We Do?
So, if the state of biotechnological and biomedical research is totally different than it was in the 1970s and 1980s, what does that mean for current funding of research initiatives? Well, it certainly does not mean that we should stop supporting basic research. Even if federal funding of basic research is not the sole source of scientific progress like Bush and the linear model suggest, it is still essential to the generation of scientific and technical knowledge that advancement and innovation is based upon. The integration of biotechnology and biomedicine in industry has increasingly tied basic research to technological development based on the profitability of the resulting applications and products. So, perhaps reconceptualizing the value of basic research based on its contributions to societal outputs rather than its funding inputs will prove a more useful framing for understanding the roots of scientific progress.
At the same time, it also does not mean that we should simply prioritize applied research over basic research. Although some critics of the linear model suggest that applied research is the remedy to the shortcomings of pure basic research, there is little compelling evidence indicating that applied research is more effective at furthering scientific progress than basic research. One study published in Science by Danielle Li, Pierre Azoulay, and Bhaven N. Sampat in April of this year examined nearly 30 years of biomedical research funded by the US National Institutes of Health (NIH), the largest non-defense research funder in the US7,8.. The study by Li and her collaborators found that both basic and applied research displayed similar rates of citations in biotechnology and drug patent applications. In other words, neither basic nor applied research appear to be better suited to actually producing new products and solutions. It stands to reason, then, that prioritizing applied research over basic would not create meaningful differences in the kinds of technological outcomes generated.
So, what should we do then? Fundamentally, we need to reevaluate how we think about basic and applied research in biotechnological and biomedical research and development. The article on biomedical research from Science also found that of all basic and applied NIH-funded research only 10% is directly cited by product patents. Even including research indirectly cited in patents (meaning patents citing research that cites NIH-funded research), that number only grows to 30%. Put another way, 70% of NIH funded research was not related to new patents, even at two degrees of separation. But if the implicit social contract between science and society is that in exchange for public financial support science will produce technologies and products that improve society, then both basic and applied biomedical research seem to be coming up short on their end of the deal. Even if this is just the rate of return on investment in research—less than a third of it generates new biomedical technologies or drugs—at the least, that is not commensurate with the justifications given for furthering public funding of biomedical research.
Rethinking How We Evaluate Research
The problem of under-delivering science is frequently framed in terms of the role and merits of basic versus applied research. Yet, research suggests that, the problem seems to be across both basic and applied research equally. This seems to be not an issue of which research type is better at producing technological solutions and scientific progress, but indicative of a broader problem (at least for NIH-funded projects) with our approach and expectations of scientific funding—the distinction between basic and applied research for funding allocation.
When it comes to the technological outputs of science, a meaningful distinction between basic and applied research does not exist. We have seen that both basic and applied research are indistinguishable when it comes to the transfer of scientific knowledge into technological solutions and products. Emphasizing one over the other or trying to determine which should receive more research funding is the wrong kind of question to be asking. The question we should be asking is: which kinds of research leads to the societal benefits and outcomes we want?
Funding for science should be based on the problem-solving effectiveness of research and its potential usefulness in society. Whether the desired outcomes of research are patents, publications, commercially-viable technologies, new companies, development of human capital, economic stimulus, or fueling knowledge-generation enterprises, the basis of scientific research to receive funding should be based on the extent to which it contributes to societal goals, not based on the whether it is basic or applied research. We need to prioritize research that lives up to the social contract of public support for technological benefits and stop rewarding research that fails to deliver.
- Though I do have a litany of concerns about the Trump administration’s approach to science policy, that is not my purpose here.
- It really is not, but that is not the main point of the argument here.
- Historian of science, Nicolas Rasmussen has written extensively and excellently about the development of the biotechnology industry in the US in his book Gene Jockeys: Life Science and the Rise of Biotech Enterprise (2014).
- One prominent counterexample is the development of CRISPR-Cas9 as a genome editing technique which has taken the field of molecular biology by storm and stands to have similar impact as many early biotechnologies. I do not want to ignore or downplay the enormous impact of this biotechnology on science and industry (my own dissertation research is focused on it!). But the development of such a potent, accessible, and widely applicable technology for biomedical and biotechnological research is the exception, not the rule, over the past several decades.
- Again, it did not.
- The fascinating story of the motivations and social context that surrounded the development of many of these drugs is outlined in Rasmussen’s book Gene Jockeys.
- Li, D., Azoulay, P., & Sampat, B. N. (2017) “The applied value of public investments in biomedical research.” Science 356: 78–81. DOI: 10.1126
- While patent citations are certainly not the only measure of the productivity of scientific research and may not capture all cases, it does provide a quantifiable estimate of the impact of research and an easily compared standard of measure to evaluate between a wide variety of disciplines and projects.