Hi Cliff, hi Glenn--
I agree with you, Cliff, about how certain situations would be classified if Census's criteria are scrupulously and consistently applied. "Are these housing units? or group quarters?"
But, here's the catch: There are gray-area or borderline situations that could go either way. And specific cases are mostly determined a priori -- during the "pre-Census" address canvassing -- based on what information Census Bureau can obtain from service-providing orgs and institutions. Early in the decennial process, Census Bureau needs to make its lists (and check it twice!) determining what addresses are in the housing units universe, and what are in the group quarters universe. And that puts the data collection into two separate lanes that are difficult to lane-change out of.
Yes, I agree with you: commercially-managed "student apartments" where occupancy in the building is restricted to roommating students might (or might not) be classified as one big group quarters facility. I expect the real determining factors are: Did the university have a roster or directory of the occupants? And did the university identify that building and share that info with 2020 Census's GQ operation before mailings and data collection began? If the answer is no, then the "student apartments" building could easily end up in the housing units lane -- even if a scrupulous reading of Census's criteria suggests group quarters designation.
Post-script: Having said all that, I will let you in on the motivation behind this fact-sheet explainer. I work in a regional planning agency that coordinates with local planners in 185 cities and townships. In the majority of our region, where urban services are provided, the region has urban (or suburban) density policies. That is: served communities need to develop with density >= X units per acre. Still... no sooner does a region implement a policy, someone will want to game or subvert the policy with creative interpretations of: what counts as a housing unit? (Also: what counts as a net acre?) So, our regional planning agency counts housing stock using the Census definition and criteria; we ask our local governments to do the same. Leaving it to Census Bureau to be the arbiter of the definition was our way of resolving the issue (and de-politicizing).
Food for thought.
cheers,
Todd Graham