Not all journals will sign the DA-RT, though some non-signatories will encourage archiving, etc. (We aren’t signing it on behalf of the Journal of Public Policy.) Not all disciplines have DA-RT-like mechanisms. Overall, my impression is that the way it’s been handled / debated in political science says more about the discipline than there being a right way to do “science”.
Here are two predictions: (1) this will reduce the volume of papers hitting the journals (because people will find the process to be a hassle); and (2) this will increase the volume of papers (because more data are now available for use in papers – the PSID/NES/GSS effect). Which will hold true?
As an aside, if data are so precious or difficult to obtain, why don’t people treat them like proprietary business information?