Vannevar Bush’s vision, while successful in many ways, has made science too focused on productivity. Restoring the ethic of discovery requires disentangling the interests of universities from those of researchers.
The Second World War was a watershed for the academic sciences. Prior to the War, academic scientists worked in a “small science” ecosystem: a loose network of a limited number of research universities, colleges, and industrial research and development labs, supported by a variety of funding sources, including philanthropic foundations, university funds, and out-of-pocket. The War mobilized the sciences to support frontline efforts. The Manhattan Project is the most well-known WWII science program, but other initiatives included Bell Labs and Western Electric.1 After the War, the political consensus favored a continuation of the generous funding,2 and a new scientific landscape was born: the “Big Science” ecosystem.3
In 1945, Vannevar Bush (effectively President Roosevelt’s science advisor) mapped out the path to Big Science in a seminal report titled Science, The Endless Frontier (STEF). Bush envisioned expanding federal funding of university and college science programs to scale up the small science ecosystem. It took five years of political wrangling before the National Science Foundation (NSF) was established in 1950 to implement this vision.4
Over the seventy years since, federal spending for academic research has doubled roughly every seven years to its present level of around $50 billion. Other government agencies have become substantial funders of academic research in addition to NSF, including the National Institutes of Health and Department of Defense. Indeed, the federal government is now the dominant source of support for academic research, accounting for about 60% of total research spending, with institutional funds contributing most of the rest.5
While Bush’s vision was successful in many ways, it did not stop at scaling up small science. Instead, it brought about a remarkable transformation in the practice and culture of science.6
In recent years, rumbles of discontent have been bubbling up about what science has become. Many are concerned that science has become too beholden to political, commercial, and military agendas,7 that it has succumbed to a bureaucratic culture of conformity and risk aversion,8 or that it is insufficiently inclusive.9
Big Science has not lived up to Bush’s vision of preserving the intellectual independence of researchers. Bush believed that a robust system of academic self-governance would defend freedom of inquiry. Instead, Big Science has largely shifted power to institutions, not scientists, and has fostered among the sciences an ethos of productivity rather than discovery. To restore the ethos of discovery to its proper place, the interests of universities will have to be disentangled from those of researchers. Only then will Bush’s vision of a scaled-up version of the discovery-driven, small science ecosystem be realized.
The Balancing Act
Vannevar Bush realized that establishing discovery-driven federal partnerships with universities required reconciling two seemingly incompatible aims: intellectual freedom (necessary for researchers) and political accountability (necessary for taxpayers).
Bush argued that scientists’ freedom of inquiry and intellectual autonomy would have to be strenuously protected: “Scientific progress on a broad front results from the free play of free intellects, working on subjects of their own choice, in the manner dictated by their curiosity for exploration of the unknown. Freedom of inquiry must be preserved under any plan for Government support of science” [emphasis mine].10 At the same time, if taxpayers are supporting basic research, their elected representatives should have a say in how the money is spent.11
In 1950, these two aims were reconciled by building the NSF as a sort of firewall, which separated decisions about research proposals from those about the funding and direction provided by Congress and the executive branch. Peer validation, insulated from political pressure, made the NSF accountable to scientists, while Congress held the NSF accountable for implementing its expressed policy aims.12 The “NSF model” has since been adopted by other federal funding agencies that support academic research.13
However, within this original compromise lies another one: the mechanism for how research funds are awarded. This accommodation led to the eventual dissolution of Bush’s vision.
The Research Grant Schrödinger’s Cat
The NSF model for funding research grants was adapted from the government’s pre-World War II system of contract research. Under the earlier system, government funds were overwhelmingly directed to applied government aims, such as aeronautics, public health, weapons development, and agriculture.
Contracts are, by their nature, agreements to produce specific deliverable products, with specified work plans, on specified timetables, and at specified costs. Such a model works reasonably well for applied research, where the path to a deliverable product is clear (a table of material properties of alloys, for example, or a map of the distribution of an endangered species). In contrast, the NSF was intended to support basic research, where the path to a product, indeed even the product itself, was unclear, unknown, or even unknowable.
In adapting the contract model, NSF crafted research grants as a kind of Schrödinger’s cat: a contract and yet not-contract—simultaneously one thing, both things, and neither. The research grant has a very different meaning for researchers than for the universities that employ them. To scientists, research grants are licenses to advance discovery. To universities and funding agencies, research grants are contracts, with specific terms and deliverables.
Exacerbating the different meanings is how the budgets for such grants are divided into direct and indirect costs. Direct costs correspond to the “not-contract” aspect of the research grant. These support the project’s specific costs for equipment, salaries, travel, and publication. “Indirect costs” represent the “contract” aspect of the research grant. These are intended to reimburse universities for the incremental costs of fulfilling the research contract, including things like power, heat, shared facilities, and library resources. Usually, indirect costs are calculated as a fixed proportion of direct costs. Linking the two revenues so directly has created many problems for the sciences and scientists.
The incentives of universities, driven by intensifying financial pressures, is to maximize indirect costs revenues. To accomplish this accounting maneuver, universities must increase direct costs, which often places the interests of universities at odds with their researchers. If established academic traditions of self-governance balanced the power dynamic, this could be acceptable. But developments such as the weakening of tenure and the increasingly competitive academic market have shifted the power decisively to universities. As a result, the incentives for researchers now align with those of their employing institutions, to the detriment of the scientific disciplines.
The Faustian Bargain of Big Science
The small science ecosystem was largely governed by an ethic of discovery. Scientists pursued reputational rewards, which were determined by the opinion of peers who could knowledgeably judge a scientist’s record of turning up new knowledge. Under this system, a single transformative discovery could establish one’s reputation and career.14 More commonly, reputations were built upon a consistent record of risk-taking and creativity.
The Big Science ecosystem, in contrast, has cultivated an ethic of productivity, where success is quantified in terms of research dollars and other often correlated measures: numbers of papers published, students mentored, and citations garnered. Such proxy metrics have a dubious correlation to discovery.
In this shifting landscape, scientists increasingly pursue productivity-based rewards, whether they lead to discovery or not. Unintended consequences follow, including incentivizing publication of pedestrian, repetitive, or trivial results; crowd-following; and prioritizing grants as ends in themselves rather than as a means to discovery.15 The heroic image of scientist as intellectual adventurer, which Vannevar Bush sought to cultivate, has been replaced by the image of scientist as a toiler in an increasingly authoritarian scientific bureaucracy.16
The productivity ethic has dangerously eroded scientists’ willingness to take the risks that are essential to discovery.17 Risk aversion is baked into the ethic of productivity,18 which bodes ill for the scientific enterprise. A case in point: the Human Frontier Science Program (HFSP) is a multinational funding agency that strives to foster an ethic of discovery by funding open-ended and high-risk research. (Full disclosure: I was a HFSP Principal Investigator from 2012 to 2016.) By nearly all measures, HFSP has been remarkably successful at promoting cutting edge and risky research. However, success has accrued disproportionately to scientists who are already well-established. For early-career awardees, the very group in which risk-taking should be encouraged, the picture is different. Roughly half of this cohort expressed difficulties in securing academic positions, and the same proportion reported institutional indifference toward their work. More than 80% experienced difficulties and delays in publication, and roughly 35% perceived a lack of peer and institutional recognition of their high-risk research efforts.19
What causes these career pressures? The pressure likely comes from dissatisfied institutions, who have observed that high-risk research does not produce as much in indirect costs revenue. The HFSP caps indirect costs at 10% of direct costs, a rate that, in another context, was claimed to pose a grave risk to the nation’s vital interests.20 The lower revenue potential for universities explains the tepid enthusiasm and career limitations that greet early-career scientists who take on high-risk research.
This is Big Science’s Faustian Bargain: researchers secure funding and career success by reorienting their scientific inquiry toward productivity, which pushes their interests even further into alignment with those of universities and their benefactors.21
The Tyranny of Metrics
Many universities and academic scientists no doubt strive in good faith to balance discovery and productivity. However, because they rely on metrics of dubious utility and precision to measure success, even the sincerest of efforts may be thwarted. Usually, scientists’ interests are given the short end of the stick for the simple reason that universities hold all of the power in the relationship.
Productivity metrics like citation impact and grant dollars project an air of objectivity and fairness. However, they say little about the quality or importance of research.22 Measures of this type can be thought of as Precisely Calculated But Arbitrary (PCBA) metrics, and they are central to promoting an ethic of production where a culture of discovery should prevail. Rewards like tenure and other forms of career progression, as well as status within one’s field, are increasingly built around PCBA metrics.
Discovery is much more difficult to measure than productivity. An ethic of discovery relies on its own imprecise measures, which might be called Informed Frankly Subjective Opinion (IFSO) metrics: the judgements of peers. IFSO metrics, for all their flaws, were an essential part of the nexus of academic self-governance that Vannevar Bush thought would be crucial to scientists’ ability to defend their intellectual independence and integrity.
Some argue that this is the way science, big or small, has always operated, and that it is the best way to ensure science provides tangible benefits to the public.23 Commonly cited public benefits include returns on investments in new products, devices, and processes, which stimulate economic growth, as well as the expertise to solve societal problems. These benefits have been used to justify public expenditures for science since the beginning of the American republic.24 Indeed, Vannevar Bush gave them a prominent place in STEF.25
The vision of universities as hives of technical entrepreneurialism is belied both by the tepid ability of most universities to bring new products to market and by the continual demand for new programs at NSF and NIH aimed at improving the commercial value of these agencies’ science. In their various efforts to fund high-risk research, NSF and NIH in effect are acknowledging that the prevailing process systematically selects against science motivated by the discovery ethic. Otherwise, why would special programs for high-risk research be necessary?
On the university side, the increasing use of productivity metrics to judge and incentivize scientists’ performance shows that universities lack the patience to support and cultivate the slow, risky, uncertain, and often long-term process of scientific discovery. Such behavior reveals a disregard for universities’ purpose in the science ecosystem, whether it be Big Science or small science.
The implications for trust in science are troubling. The public trust that scientists presently enjoy derives largely from the widespread belief that scientists are motivated by an ethic of discovery: by dispassion in the pursuit of knowledge, by objectivity in method and judgement, and by the esteem of knowledgeable peers who hew to the same ideals.26 When science is governed by an ethic of productivity, in which the norms for scientific conduct are proprietary, authoritarian, and driven by commission,27 public trust in science is eroded.28 Scientists are no longer perceived as disinterested and trustworthy advisors on matters of public interest, but as increasingly partisan and politicized claimants on the public fisc. The COVID-19 experience has shown how fragile the public’s trust can be.
Restoring an Ethic of Discovery
Vannevar Bush’s genius in crafting STEF was to draw an essentially political aim (federalizing basic science) along in the powerful wake of an idealistic vision (enhancing discovery in the universities). The compromises made to implement Bush’s vision—NSF as a firewall, grants as contracts and yet not-contracts—manifest his vision’s inherent tensions. Over time, these tensions have profoundly reshaped the science ecosystem. In the small science ecosystem, the disparate interests and motivations of scientists, universities, and funders were recognized and carefully balanced against one another. Bush thought that this careful balance could simply be scaled up. Instead, generous funding has forced all parties into a cartel, organized around the relentless pursuit of public money.29 This has put the core values of the scientific enterprise at risk.
Cartels are maintained by weaving the interests of all parties into a single Gordian knot of collusion. To restore science to an ethic of discovery, the disparate interests of all parties—academic scientists, the universities that employ them, and the agencies, private or public, that provide the funds—must be disentangled somehow. The Gordian knot that is binding the Big Science cartel together is the current regime of indirect cost calculation and reimbursement, driven by the grant-level connection between direct and indirect costs.
There is nothing sacrosanct about the NSF model’s means of assessing indirect costs. Other countries with national research programs approach indirect cost reimbursement in various ways.30 The EU, for example, assesses indirect costs at a flat 20% of direct costs.31 South African universities estimate incremental costs based on retrospective audits of universities’ actual aggregate incremental costs for research. The result is indirect costs rates that are substantially lower than those calculated using the NSF system, which incentivizes universities to inflate their PCBA projections. At Stellenbosch University, for example, indirect cost rates are assessed at around 15% of direct cost expenditures,32 roughly one-fourth the average rate for research grants in American universities. The South African approach also disentangles incremental costs (an institutional concern) from individual research projects (the scientists’ concern).
The funding model for the Human Frontier Science Program, mentioned above, offers another approach to disentanglement. Teams of scientists present proposals for risky multidisciplinary research projects. If a grant is approved for funding, direct costs are not tied to a specific program of research, but to the multidisciplinary and international composition of the team. The teams are then granted considerable autonomy in determining how the funds are spent and distributed among the principal investigators and, more to the point, among the various institutions hosting the research.33 This redistributes the administrative pressure that would ordinarily fall on a single researcher at a single university.
A group of researchers from the Netherlands goes further still in disentangling scientists’ and universities’ interests.34 In their model, called Self-Organized Funding Allocation (SOFA), a scientist receives a block grant tied to his or her record of intellectual achievement. As in the HFSP model, the SOFA model confers considerable freedom to the scientist in determining how their allocated research monies are spent. The only expectation is that the recipient will allocate a proportion of the award to other scientists, chosen by the recipient, who offer promising opportunities for new collaborations or areas of research. Here, it is scientists themselves, not institutions or funding agencies, who determine research priorities.
Full disentanglement would divorce indirect costs recovery altogether from funding for direct costs. Perhaps scientists could apply for direct costs only, while universities could apply separately, perhaps to a different entity, to recover legitimate incremental costs of supporting a broad culture of research at their university. Such an arrangement could restore scientists to the level of partnership with their institutions envisioned in STEF.
Recalibrating the professional relationship between academic scientists and their universities could also restore scientists’ intellectual independence. Tenure, proposed by Vannevar Bush as a crucial aspect of academic scientists’ intellectual freedom, has ironically become an instrument of scientists’ subordination. Where tenure has not been eliminated entirely, it has increasingly become a reward for conformity to institutional and political interests, as measured by bogus metrics of productivity. Vannevar Bush’s ideal of scientists and universities as co-equal partners has been ground into the dust, and with it has gone scientists’ intellectual freedom and integrity. Even the illusory protection of tenure can easily be nullified by dragging an inconvenient professor before Human Resources tribunals on ginned-up accusations of DEI heterodoxy—a favorite tactic of the administrative bureaucrats who enforce conformity to DEI orthodoxy.35
What reform, then? Perhaps instead of debased tenure, a scientist could be granted something akin to shares in the university “firm.” Actually make scientists partners, in other words, similar to the partnership structures in other professional organizations, such as law firms or medical practices. Such a relationship would not only put scientists on a more equable footing compared to their universities, but it would also bind all partners—the university, one’s colleagues, funding agencies—into a compact to ensure everyone’s mutual success. In addition to restoring the power balance between scientists and institutions, this arrangement would impose tangible costs on the sordid practices that are now rampant in the academy: the academic mobbing of faculty that hold unpopular opinions or insist on following a risky ethic of discovery, the bullying by tyrannical deans or department chairs to gin up money, and the domineering machinations of unaccountable, DEI-obsessed HR bureaucracies.36
Disentangling the incentives of the different stakeholders would move us closer to Vannevar Bush’s original vision of a scaled-up version of the discovery-driven, small science ecosystem. Great science came out of the small science ecosystem, and great science would come again from its restoration.
1 Jon Gertner, The Idea Factory: Bell Labs and the Great Age of American Innovation (New York: Penguin Press, 2013).
2 Richard Rowberg, Federal R&D Funding: A Concise History (Washington, D.C.: Congressional Research Service, 1998).
3 A. Hunter Dupree, Science in the Federal Government: A History of Policies and Activities (Baltimore: Johns Hopkins University Press, 1986); and Derek J. de Solla Price, Little Science, Big Science (New York: Columbia University Press, 1963).
4 Dupree, Science; Daniel J. Kevles, “The National Science Foundation and the Debate over Postwar Research Policy, 1942-1945,” Isis 68, no. 1 (1977): 5-26.
5 Michael T. Gibbons, “Universities Report 5.7% Growth in R&D Spending in FY 2019, Reaching $84 Billion,” National Center for Science and Engineering Statistics, January 13, 2021.
6 de Solla Price, Little Science; Paul R. Sanberg et al., “Changing the Academic Culture: Valuing Patents and Commercialization toward Tenure and Career Advancement,” Proceedings of the National Academy of Sciences 111, no. 18 (2014): 6542-6547; Stephen P. Turner and Daryl E. Chubin, “The Changing Temptations of Science,” Issues in Science and Technology 36, no. 3 (Spring 2020): 40-46; Matthew Crawford, “How Science Has Been Corrupted,” UnHerd, May 3, 2021, https://unherd.com/2021/05/how-science-has-been-corrupted/; and David Kaiser, How the Hippies Saved Physics: Science, Counterculture, and the Quantum Revival (New York: W. W. Norton, 2011).
7 Kevles, “Postwar Research Policy”; and Jessica Wang, “Liberals, the Progressive Left, and the Political Economy of Postwar American Science: The National Science Foundation Debate Revisited,” Historical Studies in the Physical and Biological Sciences 26, no. 1 (1995): 139-166.
8 Turner and Chubin, “Changing Temptations”; and Kaiser, How the Hippies.
9 H. Holden Thorp, “The Frontier Is Not Endless for All,” Science 372, no. 6542 (May 2021): 547-547.
10 Vannevar Bush, Science the Endless Frontier (Washington, D.C.: Office of Scientific Research and Development, 1945).
11 Alvin M. Weinberg, “Criteria for Scientific Choice II: The Two Cultures,” Minerva 3 (1964): 3-14.
12 Kevles, “Postwar Research Policy.”
13 Rowberg, Federal R&D Funding.
14 Turner and Chubin, “Changing Temptations.”
16 Kaiser, How the Hippies.
17 Turner and Chubin, “Changing Temptations”; and Helaine E. Resnick, Keegan Sawyer, and Nancy Huddleston, Trust and Confidence at the Intersections of the Life Sciences and Society (Washington, D.C.: National Academies Press, 2015).
18 Turner and Chubin, “Changing Temptations.”
19 Review of the Human Frontier Science Program 2018, Science-Metrix, December 6, 2018, https://www.hfsp.org/node/12547#book/.
20 Ronald J. Daniels, “Knee-Capping Excellence,” Issues in Science and Technology 34, no. 2 (Winter 2018); “Daniels: Proposed Cap on Indirect Costs Could Have Devastating Effect on University Research,” John Hopkins University Hub, January 23, 2018, https://hub.jhu.edu/2018/01/23/daniels-research-funding-commentary-issues/.
21 Alvin M. Weinberg, “Criteria for Scientific Choice,” Minerva 1 (1963): 159-171.
22 Kaiser, How the Hippies; and Paul Wouters et al., “Rethinking Impact Factors: Find Better Ways to Judge a Journal,” Nature 569 (2019): 621-623.
23 Daniel Sarewitz, “Saving Science,” The New Atlantis 49 (2016): 4-40.
24 Dupree, Science.
25 Bush, Science the Endless Frontier.
26 Resnick, Sawyer, and Huddleston, Trust and Confidence.
27 Turner and Chubin, “Changing Temptations.”
28 Resnick, Sawyer, and Huddleston, Trust and Confidence; and Cary Funk, “Mixed Messages about Public Trust in Science,” Issues in Science and Technology 34, no. 1 (2017).
29 J. Scott Turner, “Science and the Decline of the American Academy,” Heritage Foundation, June 19, 2020, https://www.heritage.org/science-policy/report/science-and-the-decline-the-american-academy.
30 Janek Ratnatunga and Erwin Waldmann, “Transparent Costing: Has the Emperor Got Clothes?” Accounting Forum 34, no. 3-4 (2010): 196-210.
32 “The AAS Costs Guidelines,” Stellenbosch University, 2016, http://www.sun.ac.za/english/faculty/healthsciences/rdsd/Documents/Funding-calls/2016/AAS-cost-guidelines.pdf#search=indirect%20costs.
33 Review of the Human Frontier Science Program.
34 Jop de Vrieze, “With This New System, Scientists Never Have to Write a Grant Application Again,” Science, April 13, 2017, https://www.sciencemag.org/news/2017/04/new-system-scientists-never-have-write-grant-application-again.
35 David Acevedo, “Tracking Cancel Culture in Higher Education,” National Association of Scholars, February 18, 2022, https://www.nas.org/blogs/article/tracking-cancel-culture-in-higher-education?mc_cid=a2bc86b30c#caseslist; and Komi German and Sean Stevens, “Scholars Under Fire: The Targeting of Scholars for Ideological Reasons from 2015 to Present,” Foundation for Individual Rights in Education, 2021, https://www.thefire.org/research/publications/miscellaneous-publications/scholars-under-fire/scholars-under-fire-full-text/.
36 Sb Khoo, “Academic Mobbing: Hidden Health Hazard at Workplace,” Malaysian Family Physician 5, no. 2 (August 2010): 61-67; Eve Seguin, “Academic Mobbing, or How to Become Campus Tormentors,” University Affairs / Affaires universitaires, September 19, 2016, https://www.universityaffairs.ca/opinion/in-my-opinion/academic-mobbing-become-campus-tormentors/; and Joan Friedenberg, The Anatomy of an Academic Mobbing (Lewiston: Edwin Mellen Press, 2008).
Dr. J. Scott Turner is Director of the Diversity in the Sciences Project for the National Association of Scholars.