Duck and cover! Rant incoming! Remember: Face down, ass to the blast!Here's a fun fact: Anyone can give out any degree they like. You can do it. I've done it; heck, I'm a Ph.D. granting institution. (And I thought the diploma turned out rather attractive.) But it'll be worth approximately the price of a piece of used printer paper.
A Ph.D. in particular consists of three different things:
* An academic degree awarded by an appropriate educational institution. This is the least important part. It's worth is dependant only on the reputation of the institution for excellence in the other two bits. Saying "I took classes for a PhD" is pretty meaningless---that has nothing to do with the other two facets, and in most well-regarded programs the classes are only there to provide background that you may have missed before getting into the program, for introducing you to the more advanced parts of the topic, and for introducing you to the faculty you may want on your committee. In fact, grades are seriously inflated. If you get less than a B, you must have done something pretty terrible to fail. Killing your professor's dog terrible. Saying "I wrote a dissertation for a PhD" is almost as bad. No one cares about the dissertation, other than that you have to write it. The only requirements from the institution are that it be formatted properly and turned in to the appropriate Dean's office before the deadline. (Something that most grad students fail at at least once.)
* A program of training for becoming a professor. You should be able to lecture without embarrassing yourself. You should be able to write an academic paper without embarrassing your co-authors. And you should be able to write grant proposals without embarrassing your employing institution. (These same set of skills are required for industrial research, in a modified form, as well as for academic jobs.) In practice, this is the most important facet, although in theory it should be second.
* And finally, the only part that really counts, although it is less important in current practice: Preparation for and a demonstration of the fact that you can conduct a successful research program that adds to the totality of human knowledge. This is what a PhD is.[1] A demonstration to whom, you may ask. (An excellent question!) You demonstrate your ability to a group of your new peers: other PhDs. Specifically, a group of faculty members and possibly other PhDs from outside the institution. Think of it like getting a black belt[2] or having a professional Go rating[3]: you get it for convincing others who already have it that you're worthy. How do you demonstrate it? You pick a topic with the aid of an advisor (more or less aid; there's many styles here), do the research working with your advisor and your committee (more or less (ahem)), produce papers describing the research, go to conferences, etc., and then write a dissertation describing your research program (that no one will ever read). Finally, you defend your research, somewhat like playing king-of-the-hill: your committee and other interested parties try to poke holes in you and your research in a public forum. That's it. When you convince them, you're done.
This is why I don't have a lot of sympathy with that Area 51 guy who said he has a PhD from MIT but that a government conspiracy is covering everything up. You cannot get through that last part without leaving many, many publicly visible footprints. Certainly not from MIT.
Likewise, you can get a PhD diploma from anyone, anywhere. But if you didn't go through that process, you may have a PhD but you don't have a PhD.
Note: I'm only familiar with the process in the US, and in STEM fields. But I think it's generally applicable. YMMV.
[1] http://matt.might.net/articles/phd-school-in-pictures/
[2] It's been a long time since I practised any martial arts. Have some salt.
[3] Likewise.