Flying cars were a staple of popular futuristic accounts of our present. Rapid progress in the 19th century and in the first half of the 20th century gave reason to believe that scientists would soon discover enough secrets to make an exciting and abundant future possible.
That future is yet to materialise and seems more distant now than in the past. The future keeps slipping further and further away from the present. We seem to have largely given up on the dream of zooming across the sky in flying cars, while making sitting in traffic jams obsolete.
Flying cars have not faded from discussion altogether. They have become a staple of a different genre: inquiries to the general scientific and technological stagnation ailing our era. These retrospectives link flying cars and progress more broadly. Why don’t we have flying cars? Why has scientific and technological progress slowed down? Should we give up on the dream of flying cars? Or can we somehow yet reignite science and progress?
The idea that scientific progress has slowed down may seem incredulous at first, the absence of flying cars notwithstanding. Invention of Pfizer and Moderna’s Covid-19 vaccines has after all brought to the fore the incredible power of human imagination once again.
The novel mRNA approach used in the Covid-19 vaccines has provided a very public demonstration that science can still uncover important secrets and bring immense tangible benefits to the population at large. This triumph of science has justifiably dominated much of the public discussion for the last 15 months.
Systematic evidence, however, suggests that scientific productivity has slowed down. Scientific breakthroughs like the mRNA vaccines have been all too rare in recent decades, despite massive increases in science investments over the last century. Scientific stagnation has also contributed to the slowdown in economic growth in recent decades. Science no longer powers the economy the way it once did.
The debate over the causes of the scientific stagnation has centred on two competing explanations. We have either largely run out of good new ideas, or we are looking for new ideas in the wrong way.
The former explanation—that the secrets that are still there waiting to be discovered are now fewer and harder to find than in the past—is the more popular of the two explanations for the scientific slowdown. This explanation is also the more convenient of the two. It absolves scientists and policy makers from any responsibility over the slowdown.
This “vanishing secrets” explanation of scientific stagnation is also likely wrong.
The key problem the “vanishing secrets” explanation of scientific stagnation is that scientists of nearly every era have had the same complaint, including, for example, late 19th century physicists on the brink of relativity and quantum mechanics. Over and over again, scientists have vocally given up on the future of scientific discovery only to have been proven wrong in relatively short order.
Another problem is the self-fulfilling nature of this explanation. The more scientists adopt the belief that uncovering important secrets is now exceedingly hard, the more difficult it becomes for those rare scientists who eschew this belief to get funding for their work. Scientists who would dare to attempt to discover important new secrets would increasingly find their exploratory work deemed infeasible by their peers.
Absent any overwhelming evidence that science is now harder, we should thus not give in to scientific pessimism. We should emulate earlier generations who—to our great benefit—maintained an optimistic view of discovery and did not surrender in the face of repeated and in retrospect very foolish claims that important secrets were vanishing.
The competing explanation of scientific explanation reflects scientific optimism: there are still plenty of important secrets for us to discover if we look for them in the right way. According to this explanation, lack of progress is not driven by nature but culture—the way scientist incentives and scientific institutions are presently organized.
Incentives and institutions in science are now tilted against exploratory work in favor of incremental science. As a result, far too many scientists pursue incremental research that seeks to squeeze the remaining insights and benefits from past breakthrough discoveries.
To be sure, incremental research is also useful. But the increasing focus on incremental science has come at the expense of exploratory science that that tries out and develops newer, less-established, ideas.
The decrease in true exploration of the unknown has meant that fewer ideas now develop into transformative breakthroughs. For transformative ideas are not born as such. Ideas only develop into transformative breakthroughs if a significant community of scientists tries out, explores, and develops these novel ideas. Because most ideas ultimately fail even if given sufficient attention, such exploratory work is risky. Scientific progress has stalled because too many scientists now eschew such risky, exploratory science.
The tilt in science in favour of incremental work that builds on well-established ideas at the expense of exploratory work that tries out new ideas is due to many factors. These include the aging of scientific workforce, dominance of peer review, and the obsession with citations in science.
Consider first the impact of the demographic shift in science. The scientific workforce is now much older than in the past. It has long been observed that young scientists are more eager to try out and embrace new ideas compared to their more seasoned colleagues.
The hostility of older scientists toward new ideas is well captured by Max Planck’s observation that scientific ideas advance one funeral at a time. The price of the demographic shift in science—and the related graying of grant recipients—has been a decrease in scientists’ willingness to explore and develop new ideas in science.
Consider next the effect of peer review in science. Peer review is now the dominant allocation mechanism both in scientific funding decisions and in editorial decisions at scientific journals. Scientists today depend on other scientists’ views of their work at every stage of their career.
Ubiquitous peer review puts scientists exploring new ideas at a considerable disadvantage. New ideas are initially raw and poorly understood even by those who study them. As a result, projects that explore new ideas still at their infancy are often both poorly articulated and poorly understood by reviewers.
Another source of disadvantage for scientists engaged in exploratory science is that new ideas also threaten to upend the prevailing wisdom that the reviewers’ own careers were built on. Reviewers’ well-known hostility to new ideas is thus understandable.
The dominance of peer review has steep costs. Scientific institutions today make it far too difficult to get funding for ambitious exploratory projects that would seek to uncover important new secrets. And scientific journals are far too reluctant to publish exploratory work that might later develop into transformative discoveries. Hence, the career risks involved in pursuit with exploratory science are now high for all scientists but especially so for young scientists.
Consider now the impact the impact of citations in science. Science today has come to rely heavily on citations as the measure of success. Citations—references to a scientist’s work from the work of other scientists—have in essence become the coin of academia. The obsession with citations is most evident in popular services such as Google Scholar which reduce the worth of each scientific paper and scientist to the number of citations the work has received.
The fixation with citations in science has decreased the incentive to engage in exploration. When a scientist ventures into an emerging area of investigation, the work is unlikely to garner many citations due to the lack of other scientists working on related topics.
By contrast, as a successful idea matures, the relative certainty of making discoveries—however incremental—attracts more and more scientists to the area. Because so many are working on the idea, the work will tend to receive many citations.
Because citations do not distinguish what type of science is being pursued, they cannot account for the risks that are inherent in exploratory science. The high risk of failure is an ineradicable feature of true scientific exploration. Incremental work, by contrast, is much safer in this respect.
The focus on focus on citations in scientific performance evaluation has thus dramatically tilted incentives toward more conventional research paths. Stagnation is the result of a systematic undervaluing of failed exploratory work.
To the extent that scientific stagnation is driven by the aging of scientists, the rising dominance of peer review in funding and publishing decisions, and the citation mania, there is also good news. There are constructive remedies for these problems—we can yet reignite science.
We can fund younger scientists more. We can decrease the reliance on peer review. We can start measuring and rewarding also scientific novelty in addition to scientific influence.
First steps toward reigniting science are already being taken. The Biden Administration has given $6.5 billion for the founding of the ARPA-H funding model of biomedicine. This initiative seeks to mimic the celebrated DARPA model that has been used for decades to facilitate innovative defense-related research. ARPA-H seeks to support “high-risk exploration that could establish entirely new paradigms” by taking funding decisions away from citation-obsessed peer-reviewers who are now widely considered to be hostile to new ideas.
A number of Silicon Valley initiatives are similarly seeking to offer scientists more freedom to pursue exploratory projects that what existing scientific institutions are currently offering. Arc Institute, Arcadia Science, Fast Grants, and Altos Labs are all premised on the belief that the bottleneck for scientific progress is not the lack of funding but instead the lack of freedom to pursue risky, exploratory research paths. Scientists’ incentives have become far too tilted in favor of relatively safe, incremental research directions.
These first steps toward reigniting science are important. While the “vanishing secrets” explanation of scientific stagnation is popular, the Biden Administration’s ARPA-H project and the aforementioned Silicon Valley initiatives demonstrate that not everyone shares this pessimistic view of science. There are still many scientists, entrepreneurs and policy makers who believe that we should not give up on future without a fight—we can yet reignite science.
An important caveat to these first steps is that they only influence the incentives of a very small subset of researchers. Thus, the initiatives themselves are only an incremental step in the right direction. To truly reignite biomedical science, we must fundamentally change the career incentives of scientists to be more favourable toward novel, exploratory research paths.
One way to achieve this is to start measuring and rewarding also the novelty of scientific work. One technique for achieving this involves analyzing the text of research publications based on the words and word sequences that appear in each paper, making appropriate adjustments for synonyms. Such text analysis reveals which papers build on newer ideas and which papers build on more-established ideas.
Today bibliometric services such as Google Scholar measure the worth of every scientific paper and scientists only based on one metric: the number of citations. This metric is a useful indicator of scientific influence but does not capture what type of science the work represents: whether it is incremental science or exploratory science.
The hope is that Google Scholar and other bibliometric services will soon start reporting also measures that capture the scientific novelty of each scientific paper and the career output of a scientist. University administrators and funding agencies can then start rewarding and funding scientists also based on their willingness to explore new ideas.
Such reform would bring a seismic shift to the incentives in science—back toward exploratory work and away from incremental science. A key advantage of this reform is that it would impact the incentives of nearly everyone in science. Citations would no longer be the sole coin of academia. Scientists could be rewarded and celebrated not just for successful influential papers but also for work that dared to explore new research directions irrespective of the influence the work has garnered thus far.
It is worth emphasizing that although the productivity of science has declined in recent decades, science continues to bring tangible benefits to the population at large. By all indicators, public spending on science is still a very good investment.
Scientific and technological progress brings also an important intangible benefit to the society at large. When people benefit from discoveries made by people elsewhere who are very different from them in many ways, it helps people see these other people as their allies rather than enemies. This is an important additional reason to examine ways to reignite scientific, technological, and economic progress.
There is no guarantee that the new initiatives and proposed science reforms will reignite scientific progress and economic growth. But reforming scientist incentives and scientific institutions are our best chance to do so. We should not give up on the future—and the dream of flying cars—without a fight.
Hall, J. S., 2021, Where is My Flying Car. Stripe Press (link).
Kaiser, J, 2021, “Biden Wants $6.5 Billion for New Health Agency to Speed Up Treatments,” Science (link).
Mast, J., 2022, “Inside the Multi-Million Dollar, Silicon Valley-backed effort to reimagine how the world funds (and conducts science), Endpoints News (link).
Bhattacharya, J. and M. Packalen, 2020, Stagnation and Scientist Incentives, National Bureau of Economic Research Working Paper No. 26752 (link).
Bhattacharya, J. and M. Packalen, 2021, A Novelty Reformation to Reignite Science. Manuscript (link).