I will just share my experiences (bad ones) to throw in some perspective.
I had submitted a paper once to a top conference in computer vision which was rejected for good reasons acceptable by me. But one of the reviewers, which I guessed was a top name in that area (also had a coursera course), asked me to cite 5 of his papers in the review. All were his papers.
A few years later, when I am doing my PhD, I got a paper accepted in a top conference. This time my supervisor sent me a list of 8-10 papers which I had to cite. Since there were constraints of space, he asked me to truncate a diagram and reduce the size of images to accommodate his list. Those papers had no relation whatsoever to my work. My supervisor is an IEEE fellow and has more than $1M in grants.
Citations improve H-index. H-index is crucial to get promotions and fundings. We had a very highly talented lecturer joining our university. He got 3 ICML papers accepted as first author (a top conference in machine learning) within a year, which is a fantastic achievement, but his grant application was rejected in favour of another lecturer who had 20 papers in mediocre conferences but a much better H-index. She works closely with my supervisor-- is a part of his nexus. The lecturer left my university and joined Google.
Academia is rotten with such practices. I regret that I agreed to add those citations - should have fought back. But then, fighting back is of little consequence when the system is fundamentally broken.