In applied computer science, not citing recent literature (and in this case, "recent" really means "the last couple of years") correlates very strongly with rejection. Essentially, when I review a paper where all references are old according to the standards of the field, the reason can be either that:
- there is newer related work, and the author is not aware of it or wilfully ignores it
- there is indeed no recent related work; this is at least an indicator that the problem is either solved or deemed irrelevant by the community
- the paper has been written a long time ago and has been rejected at multiple previous attempts at publication
- the paper author is just a crank, who believes he has redefined (for instance) the concept of object-oriented programming
Points 1, 3 and 4 all warrant rejection on their own. Point 2 isand 3 are at least a strong indicator that a paper should be rejected. That being said, I cannot remember ever rejecting a paper only for failing to cite recent literature. In all cases that come to my mind, this was just a minor sidenote and the paper actually got rejected for much more fundamental issues.