[Talk-us] Differences with USA admin_level tagging

OSM Volunteer stevea steveaOSM at softworkers.com
Tue Jul 11 18:18:34 UTC 2017

I'm glad Adam brings up the topic of Gores, as I'm also unclear on how such "holes" get "punched into" larger (multi)polygons via tagging.  For example, I am "sort-of-sure" (but not positive) that in Vermont, a "gore" (or grant, location, purchase, surplus, strip...usually the result of "leftovers" from survey errors) get a tag of admin_level=4 to accurately reflect that the governmental administration happens via state-level bureaucracy.  Yet, like Adam, I also have the nagging feeling of "smells wrong," because I don't understand the mechanism by which such a "hole" is "punched into" the state like this to the exclusion of the lower-level entities (like a Town).  ("Sort-of-sure" doesn't feel good enough to me, so I seek clarification).

Yes, we use multipolygon relations to explicitly do this with outer-role polygon(s) and inner-role "hole" polygon(s).  But we are not using an explicit multipolygon relation here, we are just expecting the mechanics of OSM (like a renderer) to "figure this out."  Similar things happen, for example, with Indian Reservations "punching holes" in state and/or federal jurisdictions in places like Arizona and Oklahoma (to name but two).  So, too, perhaps with military reservations, though that is not as good an example, as I don't think we regularly tag those with admin_level=2 to explicitly assert federal/national-level jurisdiction on those, though we could or might.  An additional wrinkle in that example is that since the 1950s, the USA has taken pains to extend "full concurrent jurisdiction" on all federal enclaves such as military reservations, implying that BOTH admin_level=2 and admin_level=4 might be applied to federal enclaves such as military reservations.  Our US_admin_level wiki mentions this, but it is left unclear on the correct approach we might take.

So, in short, I am asking anybody who is able to do so to please clarify:  without using a multipolygon relation, is it correct within OSM to tag, say a very large "lower 48 states" polygon with admin_level=2 AND ALSO tag admin_level=2 on, say, a national_park inside of it (which is itself inside of another polygon called a "state," with admin_level=4)?  Does "the right thing happen" if/as we do this, for example, would a renderer know to draw this as a "hole," without using a multipolygon relation?  Are semantics applied correctly to the map such that the national_park is federal, even though it is within a state?

Guidance by knowledgable people with real answers might guide us on a number of these situations, not just "Gores" (et al) but other kinds of "hole" tagging without multipolygons.  We do strive to do this correctly!

Thank you,


Adam Franco writes:
On the "Gores" point: In Vermont, while these do not have any
administrative infrastructure and are managed by the State, they *are*
surveyed and named places with defined borders (shared with their
surrounding Towns). As such it likely makes sense to preserve them as
multipolygons each with their own name and detail tags. Since these areas
are exclusive of Town/City areas, it might make sense to give them the same
admin_level even though the mechanisms of administration are different.
They aren't States themselves, so a border=administrative,admin_level=4
smells wrong. I can't speak to the situation in Maine.

More information about the Talk-us mailing list