How important is collaboration in scientific laboratories? A new paper by Pierre Azoulay, Joshua Graff Zivin, and Jialan Wang* studied what happens to research productivity when an academic “superstar” dies while they’re still actively engaged in scientific research. A superstar is a brilliant scientist who teams up with others to collaboratively conduct research and who co-authors with other scientists.
The researchers analyzed the coauthors of 137 eminent life scientists. On average, each superstar had 73 coauthors. (That number alone is astonishing, and shows how collaborative modern science is.)
Following the death of the superstar, his or her colleagues suffer a quality-adjusted decline in productivity of 8% to 18%. The authors found that this decline was lasting. Furthermore, the closer you collaborated with the superstar, the more your productivity declined. Their conclusion:
These findings are surprisingly homogenous across a wide range of coauthor and coauthor/superstar dyad characteristics. Together, they suggest that part of the scientific field embodied in the “invisible college” of coauthors working in that area dies along with the star — that the extinction of a star represents a genuine and irreplaceable loss of human capital.
As Azoulay said:
Our interpretation is that superstars infuse their scientific field with fresh ideas. They replenish it periodically and when they die, the entire field contracts so it’s really about their ideas and the effects of losing them are fairly broad and diffused.
*In press, Quarterly Journal of Economics
The Creativity Guru 
If I’m understanding Azoulay’s quote correctly, not sure if I am since I don’t have the entire context, this seems to contradict earlier postings and readings about collaboration. For instance, attributing a eureka moment to the one person who has it vs. understanding that it is really the accumulation of many ideas that came together at a certain moment. So was it that the superstar infused the group with certain ‘individual’ fresh ideas and with the superstar’s death his or her ideas died or did the synergistic/collaborate affect of the original group die because they are no longer all together and the connecting loops have changed?
I see what you mean. The methodology of this study is fairly broad-brush quantitative so hard to know for sure, but the way the authors interpret it is that ideas come from the superstar in interaction with his/her colleagues–not from the superstar’s genius per se. So that’s the second thing you said: that the synergy of the group is what dies (or at least, is reduced).
I’m guessing that a large part of that synergy, too, is attributable to the “superstar” (with his/her vast experience in collaboration) no longer bringing process expertise to help guide the group through idea generation. It would be interesting to know what percentage of those coauthors ended up going on to become superstars themselves in the wake of a superstar’s death.