Which population to use
Many large cities have a number of different population values, including wikipedia:city, wikipedia:Urban area and wikipedia:Metropolitan area. Which should we use and should we not be able to reflect all of these different values? Possibly using population:urban, population:metropolitan in addition to population? PeterIto 15:07, 4 February 2012 (UTC)
households as unit
We've got some data on villages in Africa, but it's a count of the number households. I'm thinking "population=6 households" (like unit of measurement) would be an ok way to tag this -- Harry Wood (talk) 01:51, 10 February 2015 (UTC)
- +1. If required an estimation of the number of people could be had by using a typical households' population as a rough estimate. Warin61 (talk) 10:02, 13 May 2016 (UTC)
Urban centre OR statistical area population?
A country town may have available two (or more) population numbers - one for the urban centre of the town, another for the larger area whose residents may rely on the town for supplies, services etc. Which to use? I think the larger. For example the true 'importance' of two towns, both with the same urban population, is best determined by the total number of people reliant on it. Do remember that the population number is used to estimate the facilities a town offers, if the larger area has a substantial population then the town would probably offer the services to go with that. Warin61 (talk) 10:02, 13 May 2016 (UTC)
Population during mandatory evacuations
I think we should specify (or at least suggest) the right kind of population to put in this tag. Censuses and routine demographic estimates are fine, but I think we should avoid delving into speculation or numbers that are temporary in nature.
A few days ago, a wildfire destroyed much (but not all) of the town of Paradise, California. The authorities gave a mandatory evacuation order for the entire town, which is still in effect as of writing. In the meantime, (Visualise) changed population=* from 26,389 (the 2006 Census Bureau estimate) to 0, and (Visualise) reverted it back to 26,389. The original change to 0 had the potential effect of quite literally wiping the town off the map (up to a certain zoom level, perhaps).
Depending on your viewpoint, a population of zero could be either technically correct or rather flippant. (Flippant was my first impression, but I don't know the mapper's motivation.) The practical purpose of this tag is to help renderers and geocoders give a POI the right level of prominence, using population as a proxy for notability. The town certainly hasn't become less notable since this fire, but more to the point, these use cases seem to require a more stable number than "How many people are physically inside the town limits today?" Other than those who tragically died in the fire, most of the town's residents are still residents, even if they're temporarily living elsewhere. Typically, once an evacuation order is lifted after a wildfire in the U.S., most residents do return and many end up rebuilding.
There must be a better way to indicate that a town has been temporarily depopulated. At the same time, I realize that informal settlements won't always benefit from an official headcount or attention from demographers, so there shouldn't be a total prohibition on unofficial sources.