I have plenty of nasty things to say about academia. But after about 15 years in artificial intelligence, more than 80 publications, and having worked with and interacted with 15 or so research groups, I have never seen this practice. It is possible that I have been lucky (sample size of one, too), but it is also possible that this horrible unethical behaviour is less common than in "almost every research group or collaboration". If someone have numbers with solid empirical backing, I would be happy to see them.
No it's not. It's common to cite relevant papers from your collaborators (you are working on the same topic, after all). But it's an aberration to create a semi-formal system where you force your underlings to cite your papers in completely different fields.
Might as well through in my anacdata that I've never seen this in any of the groups I am familiar with during my time as a PhD student.