Will Funders Have The Patience to Reform Science?

Discussions about metascience are frequently not about science, but about money. From reforming grantmaking regimes to new frameworks for evaluating the impact of funding, much of the science reform agenda focuses on how to allocate dollars to good ideas faster and more effectively. 

Macroscience shares that bias, often focusing on alternative models and approaches to research funding. Just last week, I was arguing that science institutions should act as a countercyclical force, helping to provide “last shot” money that balances out the bias of researchers to fold too early on potentially high impact areas. 

Friend of Macroscience and Speculative Technologies honcho Ben Reinhardt shot back a  critique on the post: although I’d identified a problem with bad instincts in research, I had failed to wrestle with the same problem in the funding context. Funders themselves are often subject to the same perverse incentives as researchers.

He’s right, of course. If much of metascience is dependent on funding, then the zeroth order question we must wrangle with is what shapes funder behavior. We can come up with all the fanciful concepts in the world about how granting might work differently, but making them a reality will involve shifting funder incentives and motivations. 

Metascience communities frequently pitch their reform proposals in terms of the pragmatic results they can deliver. The promise is that by implementing a certain set of ideas, policymakers and institutions can unlock and accelerate progress. That’s in part a matter of doing existing research work better, but more often a matter of removing barriers to taking on entirely new programs of research.

The problem is that hard research is just that, hard. Making a genuine scientific or technological breakthrough is a rare and difficult thing, even under the best of conditions. We could implement every proposal put forth by the metascience community in the last few years and it could still take a very long time to deliver a marquee success. Of course, you might get lucky, but then demonstrating that the reforms are systematically better at delivering progress than what came before still won’t become clear for some time. 

Will funders have the requisite patience to survive this epistemic fog of war? Will they be willing to spend indefinitely through years of lackluster results from grantmaking experiments, sputtering alternative research organizations, and outright failures of various kinds? 

Properly understood, these inevitable failures are a badge of honor. They reflect the fact that reforming science is hard and that the metascience community is setting genuinely ambitious goals. But I have a feeling that funders – philanthropic, private, and government – might not see it so generously. Metascience itself is a research area, one which runs the same risks of folding too early as others. 

All this means that the reform of science cannot be motivated only by the outcomes it delivers, because those outcomes may be (very) long in coming. Instead, metascience will need to accept that its pitch to funders and the broader world will have to have a strong normative component. 

This will be uncomfortable to the committed empiricist. Normativity involves advocating that science should be reformed not because we have proof-positive evidence that the reformed science will be better. Rather, reforms are justified first and foremost because they bring funding, institutions, and research more in line with our values about how science should operate. 

To give a flavor of what these values might entail, here are a few positions that I’ll be exploring in coming posts of Macroscience that I think are important to reforming the kind of science that we have today:

• We should have a science that is actively averse to fields and disciplinary allegiance.

• We should have a science that continuously lowers the costs to amateur and mass contribution. 

• We should have a science that is consciously stewarded through a form of public administration. 

This argument is not a call to abandon a hard empirical agenda in favor of dogmatism. Normative positions should be subject to rigorous investigation and must be shaped by what we learn. It is instead a recognition that a movement rooted in a commitment to normative ideals will be more robust than one rooted in the promise of imminent change. 

It is also a recognition that some of the most important problems of metascience (“What is scientific progress?”) are not necessarily ones subject to definitive resolution by experiment. The normative element of metascience is in that sense inescapable and ignoring it does no good. Metascience communities must embrace these more subjective aspects of their work, lest they abandon the answering of these big questions to others. 

---

Thanks as always to Santi Ruiz and Caleb Watney for their comments on earlier drafts of this piece.