In the US (and other places I assume), the endpoint of your PhD is determined by two things: successful completion of your research and your advisor's approval. (You could argue that your advisor's approval is the only thing). That means that the time to completion is an unknown.
This is a such a different feeling to what students I have met from the the UK/Europe experience, that it's almost like a different degree. They typically have 3 years to complete their projects, but I have no idea what "completion" means in that context.
The other one is funding, in that American PhD students could potentially be funded indefinitely, but that's a different issue...
Certainly in the UK (based on my experience in the 90s) completion of a PhD meant getting a thesis written up, getting it approved by your supervisor and then sitting an an oral examination with your supervisor, an internal academic from the same department and an external examiner who is an expert in that field. The usual rule of thumb was that you probably should have published 3 decent papers in that time.
The "3 years" came from the length of time the standard PhD student was funded for - there was no guarantee that you would finish within the three years. Quite a lot of people did PhDs while working as Research Associates - where you were typically on a 3 year fixed contract, paid a decent salary but your focus then was the research project paying your wages - not strictly your own research. Most people I knew doing it this way took between 4 or 5 years - some as long as 7 or 8 years.
For instance, you are not guaranteed a PhD in Chemistry from Berkeley even if you go for 8 years. You don't get one until your research pans out.
This is very different from some countries where you're essentially guaranteed one after 2-4 years.