I regularly assign my undergraduate students papers with a range of expected page/word length (e.g. 10-12 pages or 2500-3000 words). Is there any sort of rule as to whether or not to count a works cited page(s0 (and its equivalent in the other citation styles) in determining if a paper meets this limit? I've discussed this with colleagues and heard varying answers.
In the scientific world, whether citations count against length limits or not varies wildly by publication, including hybrid models like AAAI which allows six pages text and up to one page of citations.
I think that which way you go depends on what you want the students to learn. Some examples that would push you one way or another:
If you want them to focus on prose, don't count citations in the length.
If you want them to learn to express complicated thoughts concisely, use a short page limit and count citations (e.g., the IEEE six-page format)
If you want them to focus on referencing, count citations and text separately.
Either way it shouldn't matter too much, because you're using a range and probably have some flexibility in how you apply your rubric, so you can adjust for common sense.
Should not count.
A six-line main text can pack more convincing arguments compared to a one-line main text. That's why it's important to make sure all students have an equal share of area or real property to build upon.
A six-line citation does not necessarily bring any advantage compared to a one-line citation. For that reason, I don't see why we should penalize students who had identified some information with more authors or a longer title; that is not how we teach them to examine the credibility of a paper.
If you're concerned about them citing too many things or citing mindlessly, you may put a limit on the number of citations like some journals do. However, I wouldn't include them in the word count.