Yeah, they are super interesting. I would like to spend more time thinking about them, but my first intuition is that you can reconcile all these results if you think that firms are doing more duplicative and market-stealing research. That would be consistent with firm-level R&D generating a profit, but no collective industry-wide increase in productivity. And I think it's also consistent with Ando's argument that R&D obsolescence is on the rise (this would be the case if rivals steal your ideas more quickly). But that's distinct from the channels I discussed above - hence why I want to spend more time thinking about it!
As a preview for a future post (I hope!), I think this paper is very cool and argues it has a way to reconcile all this data. https://jlin.org/papers/GLMR-EIS.pdf
The methodology involved here is very cool and no obvious. Related to a couple of your bullets, the twin forces of a) rank age and b) ‘inverse institutional capture’ (ie where enterprising individuals are incentivized to ‘capture’ local gains within their axis of institutional advancement) seem related and mutually reinforcing. One of the things that was most affecting about being in Ukraine earlier this month was how totally absent both those factors were: there was *tremendous* leverage given to a leadership class of people under forty, coupled with very little gravitational pull within institutions for institutions’ sake. It felt like a climate uniquely commensurate with nonlinear innovation and I’m curious if a methodology for divergence-mapping such as this one can apply to less textually-legible fields (ie method diffusion, component diffusion, etc).
What do you make of papers like Fort et. al (2026) or Ando (2025) which seem to contradict “ideas are getting harder to find” with firm-level data?
Yeah, they are super interesting. I would like to spend more time thinking about them, but my first intuition is that you can reconcile all these results if you think that firms are doing more duplicative and market-stealing research. That would be consistent with firm-level R&D generating a profit, but no collective industry-wide increase in productivity. And I think it's also consistent with Ando's argument that R&D obsolescence is on the rise (this would be the case if rivals steal your ideas more quickly). But that's distinct from the channels I discussed above - hence why I want to spend more time thinking about it!
As a preview for a future post (I hope!), I think this paper is very cool and argues it has a way to reconcile all this data. https://jlin.org/papers/GLMR-EIS.pdf
Thanks for the reference, looks very interesting!
The methodology involved here is very cool and no obvious. Related to a couple of your bullets, the twin forces of a) rank age and b) ‘inverse institutional capture’ (ie where enterprising individuals are incentivized to ‘capture’ local gains within their axis of institutional advancement) seem related and mutually reinforcing. One of the things that was most affecting about being in Ukraine earlier this month was how totally absent both those factors were: there was *tremendous* leverage given to a leadership class of people under forty, coupled with very little gravitational pull within institutions for institutions’ sake. It felt like a climate uniquely commensurate with nonlinear innovation and I’m curious if a methodology for divergence-mapping such as this one can apply to less textually-legible fields (ie method diffusion, component diffusion, etc).