Everyone in our class gets a big sib, and thus gets a box. Some boxes have better, extra material than others (study guides, helpful hints), but all boxes have the same tests.
How can you be sure that all boxes have the same tests? We had some boxes have tests dating back 2 decades, and some have zero. Were students required to hand down their own tests? It may be more consistent elsewhere, but here the variations were massive; some students didn't get anything (no box, no binders, nothing) while other students had 4-5 literal boxes. Now, I am not saying an overhwhelming amount of info is particularly helpful, but that if a student has all the tests for the past 10 years and another student has 0, that isn't particularly fair. We started to move towards a study CD so that students had similar access to material, but that didn't go over very well with faculty. Now study guides are passed around by the person that creates them (and different students do different types) via internet (which is allowed here) but tests are prohibited unless clearly allowed in writing in the syllabus (and professors are encouraged to post acceptable previous tests.)
I don't see how changing questions each year would give "esoteric" information; if you're making your test questions based off of the material presented to that year's class (because lecture material often is changed from year to year), that would be the goal. You've taught the students specific information you've prepared ahead of time, and they should be tested on that info. I agree that re-using test questions allows for memorization and things, but I feel like if you test based on concepts, there is a lot of ways to ask a question, year after year (not just pertinent to a specific year's class).
I'm not saying that it has to make the test questions increasinly esoteric, but that has happened here. An example for us was anatomy; the professors reviewed the old exams and felt she shouldn't ask anything during the practical that was a correct answer on previous exams. Each year, the exams would cover increasingly esoteric information (information that would be of special or rare interest.) Not saying I agree with the philosophy used to determine testing, but I can't change a professors philosophy on repeating questions. I don't know how it is elsewhere, i just know in some of our classes there is an immense amount of detailed info presented and only a small subset of that is tested, and it isn't necessarily the big concepts. Here, often the statement is 'test questions can come from the notes, the info in class, and/or any suggeted reading for the class.' Some of our profs have been teaching here since the school opened and there isn't a infinite number of ways to ask about the attachment of a particular extensor in a specific species (and we get some odd ones such as 'in the X, what is the origin of the muscle tagged' and the tissue presented obviously isn't of the species in question or a leg that is stripped of identifying feature (ie can't see the hip or shoulder joint, can't see the hoof) with a muscle tagged and the tissue mostly covered (touching isn't allowed) with the 'id insertion' and it's a muscle where different species have different insertions. Theoretically, these aren't impossible questions, but can be really tough in 30 seconds, and require a lot of mental back tracking to get to the correct answer. My favorite was having a professor that gave review session then nothing covered in review was on the exam, and when questions she said 'obviously I couldn't give you the answers to any test questions in the review session.'
I hate to say it, but our lecture material isn't changed much from year to year. the order may change, and more information may be added (emerging diseases, name changes, new techniques) but the core of what we learn isnt' changing much.