💡 5 Quick Questions for … economist Matt Clancy on how to accelerate scientific discovery

"Technological innovation and scientific discovery is increasingly a team sport, and those teams keep growing as knowledge workers get increasingly specialized."

Sometimes one supersmart scientist isn’t enough

Perhaps nowhere is the Faster, Please! ethos more important than scientific discovery. Not only do we need to accelerate the process, but we also need to reform the process to encourage more significant advances and greater risk-taking. There’s good reason to think that big ideas are getting harder to find and that research productivity is in long-term decline. 

So, in this environment, how do we get more game-changing innovation? It’s a question economist Matt Clancy spends a lot of time thinking about. To find out about the reforms we can make to improve our idea generation — and to get a dose of optimism about what the future holds — I sent Matt a few questions, just for paying Faster, Please! subscribers.

Matt is the senior innovation economist at the Institute for Progress and a Progress Studies Fellow at Emergent Ventures. He’s also the author of New Things Under the Sun, “a living literature review about academic research on the economics of innovation, science of science, creativity, and discovery.”

1/ Are today's scientific processes and institutions biased in favor of incremental work and against breakthroughs? If so, how can that be reformed?

I think so. Part of this is simply Max Planck’s old quip that science advances one funeral at a time. Today we actually have some research that literally shows when a major scientist passes away, new entrants bringing new ideas begin to enter the scientific niche the deceased previously occupied. So that’s probably a very old problem. But when you couple it with the aging of the scientific labor force, it might be a force for mounting conservatism. Some other recent research lends support to this idea: Older scientists tend to cite older work and favor working on older topics, for example.

Another theme I’ve emphasized in some of my own writing is that the peer review system can also create bias when competition for scientific resources like grant dollars and slots at top journals is fierce. Say you are a scientist reviewing a new proposal. If the proposal is quite close to your own work and if it’s a fantastic project, you can say so with confidence because you know the area so well. On the other hand, if you’re reviewing something that’s weird and not really close to what people currently work on, it’s hard to give it top marks because you will just have lingering uncertainty. But if we only have enough money or journal space to select work that gets top marks from peer reviewers, that’s going to bias us towards great work that is similar to the expertise of peer reviewers. That stuff tends to be incremental, by definition.

Fortunately, there are a ton of things we could try to redress these issues. We just need to actually try them and evaluate those trials in a systematic way to see which ones work. We could change the formats we use to disburse research funds, for example, by looking for projects that generate disagreement instead of consensus. We could change the kinds of people who do the selecting, for example, by creating committees entirely composed of early-career scientists. We could change the default terms of grants, for example, by making them twice as long. That’s just three ideas, but I could list a lot more. I don’t know which ones would actually move the needle, but that’s why we need to try them out and study them!

2/ In a recent newsletter, you investigated independent scientific discoveries, concluding that there's a lot of redundancy in the most important discoveries but "where the real fragility lies would seem to be among ideas that are important in retrospect but not in prospect." What implications does this have for the way we approach innovation?