Do open access data policies inhibit innovation?

There has been a great deal of attention paid recently to the idea of data sharing (Van Noorden 2014, Beardsley 2015, Nature Publishing Group2015, www.copdess.com). However, the vast majority of these arguments are in agreement and present as fait accompli the idea that data are a public good and that therefore, once published, they should become open access. In fact, although there are many good reasons for data sharing, there also are a number of cogent and coherent cases to be made against open-access policies (e.g., Fenichel and Skelly 2015). The goal of this piece is not to debate the relevance or accuracy of the points made in favor of data sharing but to elevate the discussion by pointing out key problems with open-access policies and to identify central issues that, if solved, will enhance the utility of data sharing to science and society.