Ahah!
From their wiki I see this: "Use place=suburb to identify a major area in a place=town or place=city with a distinct and recognised local name and identity. Suburbs may have uncertain boundaries, may overlap with other suburbs, and are often best mapped using a node."
at
https://wiki.openstreetmap.org/wiki/Tag:place%3DsuburbWhich suggests a certain level of subjectivity, which certainly isn't the case on the ground in South Florida or southern California. And their maps are maintained by volunteers, which suggests that maybe their contributors might not fully understand the local geopolitical specifics. (Was that non-judgemental enough?) After all, how many people know that the Inglewood neighborhood in Los Angeles (home of Randy's donuts) is an incorporated city but that the Buckhead neighborhood in Atlanta is just a neighborhood? I'll bet their database is chock full of suburb <-> city errors.
If I'm reading this correctly, a legit "suburb" within a city should be tagged with the city tag and then the suburb tag. Which would mean that a suburb without an enclosing city is probably actually a stand-alone municipality. Which would be a city for DAM or IPTC purposes.
The IPTC standard isn't very helpful (or constraining, depending on how you look at it). It defines a city circularly, like so: "Name of the city of the location shown in the image. This element is at the third level of a top- down geographical hierarchy." In the extended standard, there's no definition for "city" at all. It's just a part of the Location Shown or Location Created structure.
So, I would say you're free to take some liberty here and that Hayo Bann just nailed it......
... use the city tag if it exists and the subrub one if city doesn't exist.
Actually, I would apply the same logic to town and village tags. Check out below.
More from the OSM wiki:
"Use the place=city tag to identify the largest settlement or settlements within a territory, including national, state and provincial capitals, and other major conurbations."
and
"Use place=town to identify an important urban centre that is larger than a place=village, smaller than a place=city, and not a place=suburb. Towns normally have a good range of shops and facilities which are used by people from nearby villages."
and for both towns and cities we have versions of:
"Some mappers map cities with an area instead of using a node. Cities often do not have a verifiable outline though and different interpretations of the meaning of areas representing populated places exist - more details can be found on the place key page. This method also looses the information on the city center which is important for data users, in particular for routing. Data users should not expect the area geometries of cities to have a particular meaning."
and
"Populated places (in particular place=city, place=town, place=village, place=hamlet and place=isolated_dwelling) are usually mapped as nodes since in most cases they have a well defined centre but not a verifiable outline."
Which may be true in some countries, but in the US, cities, towns, and villages are legal entities that most certainly do have very specific, legal and verifiable outlines.
