In the course of another project, I recently ran some data on the publications of 37 laureates of the Nobel prizes in Medicine, Physics and Chemistry. The results raised eyebrows in the office: they showed that those laureates, recognized for the tremendous contribution their discoveries have made to humanity, have over the course of their careers produced knowledge that has been taken up in innovation—as measured by patent citations—more widely than the work of the average US or world scientist. While this was a “quick and dirty” case study, the results exemplify the great potential of the prizewinners’ work for producing economic returns to society.
Given the ever-increasing demands for socio-economic returns from R&D investments, budget restrictions in public R&D, and the complexity of addressing grand societal challenges such as climate change, most governmental funding agencies are now aiming to strategically orient the research they support. This trend has induced fear, especially among scientists, that it could divert attention away from basic research, which is broadly recognized as being essential to the emergence of breakthroughs that will eventually lend important economic returns to society (among other things!).
Using data on 37 laureates of the Nobel prizes in Medicine, Physics and Chemistry, awarded from 2001 to 2007, highlighted the merits of research in producing positive economic returns to society. To do this, I measured the uptake, in marketable inventions, of scientific knowledge as disclosed in peer-reviewed publications. Here are the key findings:
- Slightly more than 20% of the knowledge produced by Nobel laureates eventually inspired marketable inventions (around 15 years after publication; data not shown in below graphs).
- Approximately 14% of US papers in the natural and health sciences and engineering were eventually cited in US patents, and the figure was around 10% on the world stage (data not shown in below graphs). The rates for US and world scientists cannot be directly compared to that of Nobel laureates, however, since such comparisons would not account for differences in the relevance of research for patentable innovation across various subfields. For instance, the Nobel laureates covered in my analysis published in subfields of the natural and health sciences and engineering that are cited in patents more frequently than other subfields.
- Accounting for differences in the rate of patent citations to scientific papers across subfields and publication years, the average rate of uptake of the discoveries of Nobel laureates in patents is higher than for the average US or world scientist. This is likely due to the ground-breaking character of their work. Since the databases I have used do not offer much historical depth, the signal that I captured mostly relates to the end of the laureates’ careers—most likely after their award-winning discovery. Consequently, the signal likely underestimates how outstanding these Nobel laureates have been relative to the average US or world scientist throughout their whole careers. This is also likely true of the economic impact of their discoveries.
- As they got closer to their award—most often near the end of their careers—the average citation rate of the Nobel laureates’ publications in patents declined and approached the US and world levels just before the year of their award. The research they published right after their award received renewed attention, likely because of their Nobel prize, but the rejuvenating effect did not last.
What are the implications?
Funding bodies around the globe might then ask: “Why are Nobel laureates so successful in delivering new knowledge of high economic applicability?” It could be that, compared to the average scientist (at US or world level), they generally manage their research portfolios to respond, in a more balanced way, to the demand side of research. Alternatively, it could be that by the transformative nature of the new fundamental knowledge they produce, their research stimulates the emergence of a wider spectrum of applications with a high potential to benefit society. A mixture of the above two hypotheses is also possible across the entire population of Nobel laureates.
I think the latter hypothesis is more likely, and it would provide science managers one rationale to minimize their interventions in view of producing impacts, recognizing that research is a risky endeavour that will unfortunately not always result in direct benefits to society. If instead the former hypothesis was more likely, it would provide science managers a rationale to reallocate a greater share of the curiosity-driven research they fund towards their strategic objectives, or to require, at the very least, that funding applicants provide a description of how their research might benefit society in the longer run.
Although more research would be required to decipher which of these two hypotheses, or a mixture of the two, is right, the whole research enterprise, including the work of Nobel laureates, undoubtedly builds upon both the fundamental and applied work of previous generations of scientists. This should prompt funding bodies around the world to dig further into the above question in view of optimizing the allocation of public funds for the good of humankind. This could be done by studying the broader set of factors that are likely at play (e.g., ratio of applied to fundamental research budget, demonstrated excellence of funding applicants, novelty of research proposals, institutional affiliation of scientists, level of risk taken) and that might act as drivers or barriers to the success of research in terms not only of economic but also social and cultural impacts.
Method and results
The following is a summary of the analyses for 37 laureates of the Nobel Prize from 2001 to 2007 (about 70% of the full population) in Medicine, Physics and Chemistry. For comparative purposes with US scientists and world scientists, the publications I retained were only those published from 1996 to 2012 that appeared in journals classified in the natural and health sciences and engineering in the Scopus database. I then matched these publications to the non-patent references in the USPTO to identify those publications that have been cited in issued patents up to 2016 (data extracted May 2017), thus ensuring a minimum citation window of 5 years (publication year plus 4) for papers published in 2012. The selected time periods for Nobel laureates and publications ensured that papers published by each member of the group of 37 laureates could be captured over the same time frame in relation to the year of their respective award: from 5 years prior to 5 years after their award.
Papers published prior 1996 are not available in Scopus, which limited my ability to trace the influence of Nobel laureates on innovation back to the years where they most likely produced their award-winning discoveries. According to data produced by Parolo et al. (2014), close to or more than half of the Nobel prizes in Medicine, Physics and Chemistry were awarded at least 20 years after the ground-breaking discoveries they relate to. It is thus safe to assume that the comparisons performed in this study likely underestimate how outstanding these Nobel laureates have been relative to the average US or world scientist throughout their whole careers. To avoid biases that could result from differences in the pattern of knowledge uptake in patents across scientific subfields and years, I normalized the citation rate of papers in patents by subfield and year.
I retrieved the 37 laureates’ publications using Scopus’s unique author identifiers. By using this approach, it is very likely that some of the laureates’ publications were omitted from the analyses. Therefore, I performed the analyses at the aggregate level for all laureates. I will reproduce the analyses using a more thorough cleaning of the laureates’ publication portfolios in a future post. In total, 3,291 papers were retrieved for the 37 Nobel laureates, of which 542 were cited in US patents—89% of laureates had at least one of their papers cited in a patent. Among those laureates analyzed, many appeared—as per their publication trends—to be near the end of their career, so it is not surprising that some laureates did not have at least one paper cited in patents. Also, two laureates (not counted in the pool of 37) appearing on a single prize (the 2005 Nobel prize in Chemistry) were removed because they represented obvious outliers departing from the general trend for the remaining laureates. The breadth of potential applications of their discoveries was so remarkable that a greater share of their research outputs was taken up in innovation at a much faster pace.
Compared to US and world researchers, Nobel laureates clearly stood out as having produced transformative research with greater potential to generate economic returns to society. On average, their publications were cited 45% more frequently in US patents than the average US paper (see the position of the solid black line relative to the dashed black line in Figure 1). The gap was even higher when Nobel laureates were compared to world scientists (129%, see Figure 2). This is likely due to the ground-breaking character of their work. Recall that since the databases I used do not offer much historical depth, the signal that I captured probably underestimates how outstanding these Nobel laureates have been throughout their whole careers. This is also likely true of the economic impact of their discoveries.
As they got closer to their award—again, most often near the end of their careers—the average citation rate of the Nobel laureates’ publications in patents declined and approached the US and world levels just before the year of their award. It even decreased below the US level one to two years prior to their award. The research they published right after their award then received renewed attention, likely because of the visibility brought by their Nobel prize, but the rejuvenating effect did not last as they were back to the US norm five years after their award (Figure 1 and 2).
Figure 1. Average of relative (to US) patent citations of Nobel laureates’ papers in the natural and health sciences and engineering, as at May 2017
Figure 2. Average of relative (to world) patent citations of Nobel laureates’ papers in the natural and health sciences and engineering, as at May 2017
In my upcoming blog posts on this topic, I will validate the above findings by applying another indicator of knowledge uptake in patents that will be based on simple proportions of papers cited in patents normalized by subfield and year—for example, by building random samples of US and world papers to replicate the distribution of the Nobel laureates’ papers across subfield and year. Because most papers are never cited in patents, normalizing the number of citations of a paper by the average in its subfield and year can lead to extreme outliers even with just a few citations. In turn, these outliers could easily inflate the average of a group of scientists like the Nobel prize laureates. Following preliminary checks, there do not seem to be problematic data points in the pool of papers by the Nobel laureates analyzed here, and I am therefore confident in the results. In the future, I will also explore in greater detail the positive outliers among Nobel laureates to give an indication of one dimension that might help in detecting future Nobel prizes. I will also explore the citation impact of Nobel laureates within the scientific literature—a hint: it is VERY high.
Note: All views expressed are those of the individual author and are not necessarily those of Science-Metrix or 1science.