Part of the issue must be that data analysis can largely be complete drudgery, but I also buy into the notion that statisticians may be increasingly in high demand as Hal Varian, quoted as the chief economists at Google, seems to believe. We have certainly seen a revolution in the availability and breadth of public microdata, and Google has a boatload more stuff, public and not. But grad students should probably still keep in mind that the trick is to have the vision in addition to the skills: be able to develop the ideas that specify what questions to ask with the data, not just be able to use the data convincingly.
I've probably just self-identified with that remark, but so be it!
