According to a recent NYT article (available here):
For those who attempt it, the doctoral dissertation can loom on the horizon like Everest, gleaming invitingly as a challenge but often turning into a masochistic exercise once the ascent is begun. The average student takes 8.2 years to get a Ph.D.; in education, that figure surpasses 13 years. Fifty percent of students drop out along the way, with dissertations the major stumbling block. At commencement, the typical doctoral holder is 33, an age when peers are well along in their professions, and 12 percent of graduates are saddled with more than $50,000 in debt.
These statistics are revealing and misleading at the same time. On the one hand, they do give some indication of how long and arduous the journey is to obtain a PhD. There are no shortcuts. But on the other hand, these statistics clump all disciplines together (sciences, engineering, humanities, etc.) and so don't provide a good estimate of how long it usually takes in a particular discipline. It does seem that science and engineering doctoral students take less time than humanities students (largely, I suspect, due to greater funding, better job prospects, and fewer teaching commitments). So this figure is an overestimate if you're thinking about science and engineering. In addition, the figure clumps in all types of educational programs - public, private and every hybrid in between. Since the amount of funding one gets greatly influences the speed (and attrition) of one's doctoral studies, this clumping may also be inaccurate.
I can only speak for my experiences at Berkeley, but I am familiar with a variety of factors that can lengthen one's time in graduate school. For me, the most important issue is funding. If I were guaranteed more funding (as, according to the article, is routinely done at Princeton), there would be less of a need to be a teaching assistant (GSIs as they're known here) and so could spend more time on coursework and research, theoretically finishing faster. But I happen to enjoy teaching so I'm not sure I would wholeheartedly welcome the elimination of teaching, in exchange for better funding. But since teaching here is a half-time job, it essentially means that on average I probably work about as half as much on my research as a peer at another institution who doesn't need to worry about teaching for money.