A long running debate in the modeling of innovation is whether discoveries are sure to arise by targeted research investment or by unexpected moments of invention. Suzanne Scotchmer lays out the different models in her book Innovation and Incentives. (Side note: I think every patent scholar should read this book. I would put it on my list of classic patent scholarship, but it's not pre-2000.) I tend to fall in the middle, as I usually do - some advances can be targeted, and some must remain unexpected. I think it is hard to justify a fully path directed model or else, as Scotchmer noted, everything would have been invented one or two hundred years ago. Then again, sometimes things are invented and the world is just not ready for them.
This brings us to serendipity, or rather Serendipity. Serendipity, by Bhaven Sampat (Columbia School of Health) came across my SSRN feed this morning, and I was all ready to write about it when I saw that Mark McKenna (Notre Dame) beat me to it, at Jotwell.
Professor Sampat goes about testing the role of unexpected discovery in pharmaceutical advances. It's an impressive study with impressive results, finding that money spent targeting one disease may often lead to results for other diseases. It also shows that funding in basic science can lead to specific but unexpected results; that's good, because I'm a fan of funding basic science.
I'll say little else other than to point readers to Professor McKenna's summary and analysis, which is thorough and insightful. I only have one addition those remarks. The jotwell discussion focuses on government policy for grants - targeting, amount, gaming, and so forth. I would add that the findings, if we believe them truly to be about serendipity, hold up outside of grant funding and apply to general inventive efforts. So long as there is some incentive mechanism, patents, prizes, grants, etc. that gets people doing research then we might hope to see unexpected spillovers from those research efforts.