Why did it take a private foundation to do public science right?

Microsoft co-founder Paul Allen funded the Allen Brain Observatory, a detailed, rich data-set derived from parts of a mouse-brain: what's striking is that the Allen Institute released all the data into the public domain, at once, as soon as it was available, which is exactly what you'd want the publicly funded alternatives to do, and what they almost never do.

Mark Humphries discusses how the incentives for public science — the need to publish a lot of papers, quickly and early, in order to secure grants — means that publicly funded scientists end up hoarding their data, lest they inadvertently give someone else the material that they need to work on their next paper. By contrast, the Allen Institute's targets are "tight deadlines for reaching project milestones, rigour of the methods, and quality of the resulting science."

Humphries argues that universities can get off the publish-early/often treadmill and its data-hoarding demands by reapportioning their own budgets: less on new buildings and administrators, more on long-term investments in high quality science and development of scientific tools like high performance computing software and analysis frameworks.

What all these have in common is their origin in dedicated, privately funded research institutes. These researchers are somewhat immune to the science incentive problem that pervades universities. This is because universities drive the quest for money. Research grants pay a lot towards universities' infrastructure, services, and administrative people. So universities want grants. And papers, as noted above, play a key role in getting grants: so they want papers too. (In the UK we also have the direct equation that papers = money, thanks to the REF).

A solution is thus that universities should adopt the private institute model: stop pressurising researchers to obtain grants and papers. Instead they could spend their own money sustaining the research programmes of their own researchers (rather than on, say, yet more bloody buildings , or administrators). This would remove the pressure to get short term grants, but leave open the need for high value grants for major programmes of work. Reward quality and rigour, not output. Reward the work effort, not the luck of the draw in where the paper finally came out.

How a happy moment for neuroscience is a sad moment for science [Mark Humphries/Medium]