international - Why is there such variation in PhD length internationally?

I'm an undergrad in Australia, and at this stage I want to go on to do further study in maths (or maybe physics/CS, still have a few years till then).

Should I do so, I imagine that I will want to complete a PhD overseas. Possibly at a university in the US or maybe Oxbridge in the UK.

In Australia, PhDs typically take 3 or 4 years, and I would be eligible to go straight into a PhD upon completion of my undergraduate degree (it is research focused and includes an honours year, see here).

But, much of what I have found for overseas institutions has been in the range of 5-7 years, or maybe a 3-4 year PhD with a few years doing a master's degree beforehand.

Does this indicate that the level of rigour and amount of work produced is significantly different for various PhD programs, or is it the case that some programs are more relaxed and simply spread the same amount of work over a longer period of time?

Answer

In addition to @aeismail's points, in the U.S. the undergrad degree (B.S. or B.A.) in math is typically rather thin, due to "breadth" requirements, so a year or two of a PhD program is spent catching up, in comparison to most other educational systems in which specialization occurs earlier (and perhaps high school math education is more intense).

Some decades ago, it was the style in elite places in the U.S. to have people finish a PhD degree as quickly as possible, often in three years, as proof of ... something. This was plausible under the hypothesis that students at such places had an unusually good background.

In fact, given the way professional mathematics has evolved, spending more time learning things and maturing before hitting the job market may be wise. In any case, no one is creating artificial obstacles to any student's quick graduation! In my current institution, there really are no "required courses", in the sense that there are some modest proficiency exams in standard material that need to be passed, and courses help prepare for those, or can substitute for proficiency exams to some extent. Thus, a well-prepared student can "test out" of requirements.

One underlying problem seems to be that people take as much time as is allowed, so if it is understood that one may take six years "if necessary", then most people plan to use up that time. Not that they're "forced to" or "kept from graduating earlier". And then there's the reasonable fear of facing the job market that leads to "avoidance".

I would claim that "having to teach" is not a serious impediment to quick graduation. However, its relatively immediate gratification can seduce people away from the far-less-immediate gratification of research and study.