Science and Technology

This post provides a quick overview of claim articles in New Things Under the Sun related to the interplay of science and technology.

Science is good at making useful knowledge

We’re worried scientific papers get prestige for arbitrary characteristics valued by ivory tower, but uncorrelated with truth and usefulness
Some view academic economics as prioritizing adherence to free market ideology; in fact, citations from economists also predicts citation by other social sciences
Use of science by inventors also suggests science is doing useful stuff
- Patents increasingly cite scientific articles, and these citations seem to capture some form of actual use of the ideas
- Patents building more closely on science seem to be more valuable
- Citations from scientists not only predict citation by patents, but also predict how valuable the citing patent is
(go to article)

Science seems to be particularly helpful with invention when…
- …an inventor is changing fields
- …technologies lie very far from what’s normal
- …the technological domain is very finicky: small changes or tweaks can have big (potentially bad) effects.
A scientific basis can also help communicate to others that an invention is valuable
(go to article)

Several natural experiments show more science leads to more innovation
- WWI significantly disrupted science and fewer new scientific words showed up in subsequent patents
- US science funding was shaken up at the end of the Cold War; technologies reliant on science that got more funding began making more use of science, and vice versa
- University funding windfalls from better-than-expected football seasons leads to more science, more patents, and more patent licensing revenue
- Plausibly random variation in grant support from the NIH for specific scientific fields, is associated with more papers and more biomedical patents in those fields
Funding for basic science seems to be correlated with the productivity of relevant industries, with a multi-decade lag
(go to article)

Most inventions do not seem to directly rely on science
Many more are linked to science via some chain of citation though (they cite a patent that cites a patent that… cites a scientific article)
Evidence suggests a surge in patenting in one class of technology tends to predict a later surge in “downstream” technologies
Technologies that do not usually cite science directly tend to lie downstream of technologies that do
Science may have larger indirect effects on technological progress
(go to article)

A good rule of thumb is that there is a 20-year gap between science and technology
One line of evidence: correlations between science funding and subsequent productivity growth
- The best statistical fit between the number of journal articles in some field and TFP growth in manufacturing industries reliant on that field is about 20 years
- More sophisticated Bayesian methods also find a 20-year fit is most likely, using data from agriculture
Another line of evidence: patent citations to science
- The average gap between the date a scientific paper is published and patent citing it is applied for is 17 years
- Most patents don’t cite papers, but most are indirectly linked to patents that do.
- The average (shortest) gap between a patent and a paper is just 7 years, allowing for these indirect citation trails
(go to article)