Reading material in my notional in-tray includes undergraduate project write-ups, drafts of thesis chapters or preliminary results from PhD students, a conference submission that I agreed to sub-referee, and some older papers that I should cite as background work in new papers. Also a journal paper and a conference paper that I working on, with co-authors. It does not currently include journal papers that need referee reports (hooray) or grant applications in need of referee reports (double hooray).
Now, I ought to happily lap up at least some of the above material, but I reckon that I am not a reader by nature, despite my job title. Or at any rate, not the perfect academic. When I actively work on a problem, it gives me the incentive to read up on the ideas and techniques that have been applied to it in the past. Contrary to the "natural" order of things, for me, study does not come before research.
Some thought on different types of reading material: journal paper submissions to review are rarely interesting because they are yesterday's news; conference submissions have the advantage of being new material being released for the first time. Textbooks - I have tried to read these, and never get very far. Reading textbooks feels like too much a passive activity, plus, they've got the "yesterday's news" problem. Research proposals may be interesting, if only for all the wrong reasons. You get to find out the authors' salaries, which is one of the many things that is wrong with the process of evaluating grant applications. Trashy novels - by far the best bet.