Why do people say you shouldn't do a PhD unless you want a career in research?

Research is not just the most important aspect of any PhD program worth the name; it's the entire point. Everything you do as part of the program (teaching duties aside) is to prepare you for a career as a researcher. So getting a PhD without any intention of putting this research training to use would be like spending years as a baking apprentice without any intention of becoming a baker: maybe not a complete waste if you really like baking (but not enough to make it a career?), but almost inarguably not the best use of your time.

Lots of people get PhDs and don't become researchers, but if these people had known beforehand what industry they would end up in, odds are they could have gotten a better jump start on their career by getting a Master's degree focused on that industry and entering the workforce after one or two years (rather than 5+ for the PhD). There are exceptions to this principle, but they are relatively uncommon, and usually involve industry jobs that somewhat resemble research anyway.

I can think of only two other justifications for getting a PhD: learning for learning's sake, and prestige. The former is a noble goal, but if all you want to do is learn without contributing to the body of knowledge in your field, this is contrary to the spirit of (good) PhD programs as stated above. Furthermore, you can learn all you want in your free time without entering a PhD program. Find out what the standard graduate-level textbooks are in your field, and work through them. If your knowledge gets up to the research level in a particular topic, you can start reading recent research articles. (Your local library will likely have access.) As for prestige, this may be subjective, but spending five years of your life for a piece of paper is unlikely to be satisfying in the long run. And, speaking as someone who has a PhD, you'll soon learn that most people will not really want to address you as "Doctor."

Having said all that, if you do like (or think you would like) research, but are put off solely by the expected competition and stress, I urge you to be more open-minded. Academia is certainly competitive and stressful, but so are most careers worth having.

Although the primary goal of a PhD is to train you in performing scientific research, a career in academia is definitely not the only possible result. In fact, more PhDs end up not staying in academia and not becoming a professor. A lot of people I did my PhD with either work for research institutes or consultancy companies. They are in positions where their abstract thinking skills serve them well, and are appreciated by their employers. I ended up in a data scientist like position, where my programming skills, abstract thinking, quickly learning complex topics, and data analysis skills really shine. In this field a PhD is not exactly a requirement, but a big plus nonetheless.

Where your PhD hurts you if you for example apply for pure programming roles. There you have to compete with people who have exclusively spent their time programming, and your research skills are not a plus. So, it is also a matter of finding the right position where a PhD adds value.

Note that this answer may depend on the country you live. My impression is that in the Netherlands the attitude towards PhDs is not very negative. The impression I get from this site is that this is less the case in the US.

Why do people say you shouldn't do a PhD unless you want a career in research?

Wow, do people really say that? That's the most foolish advice I've ever heard.

An active researcher at a university produces some significant number of PhD students. There is simply no way for the number of research jobs to grow fast enough for all of those students to get research jobs. This is why people talk about PhD birth control.

Better advice would be:

1. Don't do a PhD unless you think you'll enjoy doing research in the process of getting the PhD.
2. If you have your heart set on a career in research, readjust your expectations and figure out a backup plan so you won't be bitter and unhappy in the overwhelmingly likely event that it doesn't work out that way.

I think doing a PhD in a subject you like for the love of the subject/topic is a completely valid reason for doing one. Whilst a PhD is primarily structured with an academic career in mind, you do learn very useful transferable skills that are of great value to the industry-problem solving, quick to learn, logical thinking, etc. A career in academia may not be for everyone who completes a PhD and depending on the subject, there are arguably better prospects in industry with better financial remuneration.
As mentioned, there are lots of jobs in industry that specifically require a PhD qualification-quantitative analyst, engineering, etc (see siam).

So in most cases, whether you remain in academia or not, a PhD qualification will still serve you well. I do however think this is very much subject dependent.

People say it as a warning, not as a truth. Many bright young people, who love knowledge and discovery and love immersing themselves in a subject (even if the work is challenging), do a PhD because it feels like a dream job. But afterwards, they struggle to survive in the competitive research environment, and struggle to find a decent-paying job elsewhere. If anything it seems like the PhD set them back.

Nobody wants to see yet another young adult waste half a dozen years of their life. Hence this warning.

However, not all students are identical, and not all PhDs are created equal. The years I spent pursuing an engineering PhD (which I did not complete) were among the most valuable years in my life. Since I created my own highly innovative research topic, I quickly got a huge amount of experience in both pursuing, and communicating, controversial ideas. I learned how to manage the egos of the establishment, the critics who will shoot down anything that doesn't fit well with conventional ways of thinking, and how to express a complex new approach in a thirty-second or 2-page limit.

Now, I'm an entrepreneur and innovator, following my dream of bringing about real change in the world, and these skills are vital to me!

The warning is, you won't automatically get that out of a PhD. But if you're confident that you can make it happen, who is anyone else to say you can't?

You seem to have the answer already. PhD programs are training programs for research. We generally don't waste time and money training firefighters or cops that we don't deploy into service. The same reasons hold for PhD students. If you don't want to be a professional researcher, a PhD program is pretty inappropriate, a lot of work, and expensive for the funders.

Although there are many different contexts, depending on subject matter, I really don't think that there's any general understanding that getting a Ph.D. is "exactly" "preparation for research". For that matter, there does seem to be a general misunderstanding of the various possibilities for "research", given the noise about "impact factors", and the belief system in which a clever person doesn't need to know much to "do research", and so on. (True, at the other end are the gate-keepers, status-protectors...)

In my field, mathematics, I think the idea is that engaging seriously with live mathematics, with a few years' preparation, is the most plausible way to develop a sense of what it really is... whatever one does subsequently. Being a teacher of mathematics is arguably enhanced by having some idea of what the subject is, although, indeed, this does not automatically make anyone a good teacher. My students who have "gone into industry" generally report that everything else is easier than was their PhD work. Perhaps a good PhD extends one's horizons, one's self-knowledge. Gosh! :)

There is also that "scholarship" thing, ... Organizing and preserving human understanding?

I myself don't find the start-a-startup or bring-a-product-to-market version of "research "very interesting", and that helps explain why I'm not rich. :)

If one is interested in something-or-other, while it is in-principle possible to learn much of what's known on one's own, on-line, whatever, in fact there are some senior people whose expertise is not made obsolete, nor eclipsed, by on-line stuff. Then it's useful to be in close contact with them, if the goal is getting a sense of the extreme cutting-edge of the thing, whether it's "products" or understanding of very old things.

(Also, I note that the objection that doing a PhD with no plans to "do research" in a conventional sense is somehow "wasting resources" overlooks the point that nearly all RAs and TAs are paid very low wages, hence, are a terrific bargain...)

In fact, I do claim that the miracle of "getting a PhD" is that if one has no serious financial obligations, that point in one's life is a marvellous opportunity to behave very idealistically... and worry about "making a living" later. (Yes, if the "opportunity cost" bothers you, then don't do a PhD.) That is, people with the luck to have the opportunity to be "pure intellectuals" for a few years when they're young and idealistic and energetic do indeed have good fortune. Worrying about optimizing job situations can wait, maybe?

To take a different tack than some of the other answers: getting a PhD typically involves spending several years working your butt off for fairly low pay (or, in some fields, no pay, or even negative pay in that you may take on debt). That's in addition to other more minor negatives you may or may not encounter depending on the personalities you interact with; you may be treated as slave labor, browbeaten, ignored, or looked down on. For many people, the light at the end of the tunnel is the possibility of a career where they get to do research as a living. If that's the career you want, a PhD is usually required. If you don't want that, a PhD is a lot of work with limited payoff.

In short, the main difference between having a PhD and not having a PhD is that if you have a PhD you can be a professional researcher. If you don't want to do that, you probably don't need a PhD, so why go to the trouble of getting one?

That said, I personally don't agree that you should never get a PhD unless you want a career as a researcher. If you're in a good program, getting a PhD can be a fun and stimulating experience (although still a lot of work). The main thing I would say is, you should be deriving some net benefit from getting a PhD. You can benefit either from the experience of getting it, or from the doors it opens for you after you get it, but if what you're getting in those two areas isn't worth the effort you're putting in, you shouldn't do it.

In Sweden you have the possibility to become a industrial doctoral student. This might be a good compromize if you like to research but want to end up in industry and in an R&D department on a (larger) company.

Example from Chalmers University of Technology: http://www.chalmers.se/en/research/doctoral-programmes/becoming-a-doctoral-student/Pages/industrial-doctoral-students.aspx

I have worked as a researcher for 20 years and studied part-time courses in business and information analysis but never did do a PhD. A colleague of mine who did do a 7 year PhD following a BA and MA said the reason he did it was to stay in academia, however he felt like that much work involved in carrying out research even though his subject was psychology would be better spent applying the techniques and tools that he had learned to work as a researcher, running large postal surveys, analysing statistical data, managing research projects etc. I realised that he knew SPSS inside out, could explain statistical terms and concepts to me that I couldn't understand and teach me certain research methods that I hadn't heard about. So with all of those advanced skills, I'd agree that research work would be a job that wouldn't be too difficult to get following University.

I saw an article today on my Linkedin Pulse titled: 24 Highpaying jobs for people who hate stress

At least half of them requires master to PhD level education. So I'd say that's not true, but it depends on what your PhD is on.

Also, I'd like to point out an article from Hootsuite yesterday (though this may only apply to the tech industry for now). The article claims that traditionally, the only way to climb up the corporate staircase is to slowly move up the manager position( and then manage more... a team, a department, a company, etc.) However, they claim that companies like google (and themselves) long recognized that some highly skilled people are not fit to be managers and don't want to manage other people, so they set up alternative career track "the guru track". This basically allows people with high knowledge in a specific topic to move up, without having to manage. I don't know if this is the path the industry (and any other industry) is heading, but if this becomes more common, I suspect that graduate education will be more valued than they are now in the corporate world.