My paper "To Explain or To Predict?" has just been accepted to the journal Statistical Science. The working paper with the most recent version is available on SSRN.
The process of writing of this paper (and its various earlier versions) has been an extraordinary experience, with feedback from many colleagues in many disciplines and across the globe. I thank all of them for their invaluable comments and the lively discussions. Hopefully the paper will raise awareness to the important distinction between modeling for explanatory and predictive purposes, and will lead to the integration of predictive analytics in many disciplines that are currently monopolized by explanatory modeling.