Open Data should mean CC0, not CC-BY

Several months ago, NIGMS released a RFI on The Need for and Support of Research Resources for the Biomedical Research Community. Our group has quite a bit of experience building information-based research resources (through grants generously awarded by NIGMS). So based on our recent experiences integrating data for inclusion in Wikidata, Ben and I took this opportunity to formulate some thoughts around data licensing and how they affect data reuse.

The full text of our response is embedded below (with input and edits from co-signers). Most notably, we strongly advocate that NIH-funded information-based research resources should be released via a CC0 / Public Domain license. We also argue that despite the superficial reasonableness of the noncommercial provisions of CC-BY-NC and even the attribution requirements of CC-BY, both have the effect (usually unintended) of stifling data reuse. An this seems particularly counterproductive at a time when the NIH seeks to promote the principles of FAIR access to data.

While I think the entire text is worth a read, I also took the liberty of highlighting two passages that I feel most strongly illustrate both the problem and the opportunity that licensing choices provide.

Relevant and timely discussions are also happening here, here, and here. UPDATE: Also see this Nature News article from this morning.