When should a legally-designated city in Alberta be a city or town? Should that be based on population (and whether it’s part of an urban conglomeration) or its current status according to Alberta Municipal Affairs? Currently, there are two cities tagged as town, Beaumont and Chestermere, both of which have been elevated to city status withiin the last 11 years (in 2014 and 2021 respectively) and are part of the Edmonton Metropolitan Region (EMR) and the Calgary Metropolitan Region (CMR) respectively, but have populations below 30,000.
The legal status should go on the border_type tag on boundary relations, i.e. border_type=city for legally designated cities. For example, we have an example in the US of a legally designated city with zero population 
.
The place tag is used to tag the relative importance of population centers, and I’ll leave it to Canadian mappers on where to draw that line. It’s a challenging topic.
1 Like
Well, the province-level OSM Wiki page hasn’t provided any rough guidelines on this, only a list of places but not any info for appropriate place tag. Some Alberta mappers have insisted on 1:1 correlation of the place tag with the legal (AB Municipal Affairs) designation of hamlet, village, town and city (as well as AB-specific place types such as summer village and urban service area, which are equalled to villages and cities respectively). Some insist this should be based on population as described in the wiki (though that has its downsides especially on sparsely populated areas, which includes most of Alberta).
To begin with the existing town tag usage, most instances in Alberta relate to their present designation. As for the cases of Beaumont and Chestermere, they were retagged to cities when they became cities, but were reverted to towns based on population of the time (in 2016 federal census figures from Stats Canada).
(To be honest, I’ve been living in Alberta for over 7 years, and has been mapping there since. I was also on the Talk-ca mailing list though it was much quieter and most discussions are now in the Community Forum).
Curiously, smaller cities of similar population as Beaumont and Chestermere such as Wetaskiwin and Brooks are tagged city. The choice to restore town tagging to Beaumont and Chestermere has to do with them belonging to a metropolitan area.
I would consider restoring city tagging for the two cities in question, but there should be like a standard for AB. Either it must tie with the official designation by Municipal Affairs, or use a population-based criteria that fits well for AB.
Are these the only possibilities? If Beaumont and Chestermere have been demoted to town based on their location within a metropolitan area, that heuristic could be applied consistently throughout the province if desired.
That was all the possibities I can think of basing on existing practice I observed, but another reason I can infer is some of the other smaller cities retained their city tagging because they are the most important settlement in their surrounding area, which is sparsely populated.
For background, the mininum population for an Alberta urban settlement to become a city is 10,000. For comparison, the suggested minimum population for a settlement to be tagged place=city from the wiki is 20,000. As for Beaumont and Chestermere, they were retagged to city when they became cities themselves in 2015 and 2021 respectively, but were changed back to town based on their tagged population at the time they gained city status. Actually, from their 2021 federal census figures, these can now qualify for city tagging from the present general guideline based on population (Beaumont now has about 20,000 and Chestermere 22,000).
The reality on the ground is that Beaumont, Chestermere, and Wetaskiwin are cities. The Government of Alberta does not care that OpenStreetMap might think that settlements with a population below 20,000 are not cities - it will treat Wetaskiwin (pop. ~12,600) as a city. We have to go with whatever the higher level of government says because the alternative is to create an OSM-specific system of settlement types and apply it worldwide. Doing that would a) be a nightmare to keep up to date, and b) lose useful information about the names and heirarchy of settlement types in every country and their subdivisions. Eg. List of terms for administrative divisions - Wikipedia Using the local designations, whatever the rules - if any - that locality might use, at least keeps OSM consistent with reality, even if that is inconsistent from country to country.
Well I have long assumed long-standing Alberta tagging practice sticks to the Municipal Affairs designations for what place type an urban municipality gets, but we got people here such as @Viajero_Perdido which insists on using population-based criteria as found on the wiki and the bar is higher if the city happens to form a metropolitan area (as with the case of Beaumont and Chestermere).
Oh hi. Yes, that was me, in a minor edit war over Chestermere, and I think I demoted Beaumont as well. My reasoning:
- OSM rules (the wiki) override outside-of-OSM rules (politicians’ delusions of grandeur).
- How can a renderer tell Calgary is more important than Chestermere, if there’s only room for one label? Do we expect renderers to dig deeper and compare populations, etc? It bugged me to see, on our favourite renderer, the whole Calgary region labelled as Chestermere. (I’d call the fix tagging-to-rule, not tagging-for-the-renderer.)
- In a similar vein, we, and British Columbia especially, have many mighty rivers humbly named, eg, Horsethief Creek. No way am I jumping across that, so, per the wiki again, it’s a true river.
Cheers, VP.
I think the official designations are informative, and generally they accurately represent the level of importance of a given municipality relative to others in the province, but slavishly sticking to the official designations misrepresents some of these places’ actual importance. There are greater political motivations beyond population counts for municipalities to choose to incorporate as a city vs. town vs. village, or not incorporate at all and stay a hamlet.
With respect to Chestermere and its place in the greater Calgary area, to me calling it a place=city when it’s actually the fourth largest of the satellite municipalities around Calgary is silly: Airdrie (place=city) is far and away the biggest (~75,000) and Cochrane and Okotoks (place=town)—officially towns—are larger than Chestermere. I would also submit that of these four municipalities Chestermere is easily the least important: it essentially has no commerce and industry of its own. It is in the truest sense of the term a “bedroom community” of Calgary: people live in Chestermere, but they work in Calgary. It has no underlying reason for its existence other than to serve as a bedroom community to Calgary. To tag it as place=city and give it the same importance as Calgary is absurd to me. I won’t say this is definitively why Chestermere wanted to incorporate as a city in the first place, but anecdotally a lot of public perception was that they wanted to incorporate as a city mostly to obstruct Calgary from eventually annexing it (as Calgary has done to many former bedroom communities in the past, like Bowness, Forest Lawn and Midnapore).
What do you do about municipalities that could apply to become cities under the MGA (Municipal Government Act of Alberta), but choose not to? The aforementioned Cochrane and Okotoks are more than twice the population of Wetaskiwin, Brooks, Cold Lake and Lacombe—all tagged place=city. Other towns such as Blackfalds, Strathmore, Sylvan Lake, Stony Plain and Canmore have all exceeded the MGA’s 10,000-person threshold, and chosen to remain towns. (And they’re tagged as such in OSM.) Blackfalds has almost the same population as Lacombe, but the latter is a city and the former is a town; is this really an important or germane distinction to be making when the population of each of these municipalities is only about a tenth of nearby Red Deer?
We already paper over the nonsensical official designations of Sherwood Park and Fort McMurray as “urban service areas with specialized municipalities”—effectively ‘hamlets’ insofar as they don’t have their own municipal governments separate from the surrounding rural area—and have tagged these both as place=city regardless.
I think we have to make some concessions to the reality of many ‘cities’ being satellite towns of Edmonton and Calgary. Maybe if there was a greater “big city” tag in OSM it wouldn’t be an issue, but to tag Chestermere, Airdrie, Sherwood Park and the like co-equals to Edmonton and Calgary is nonsense.
3 Likes
That’s the argument for border_type=*, which was mentioned earlier. By design, the key intentionally hews to official designations, regardless of any statistic or observable characteristic. This is even more obvious on features that represent political subdivisions but not populated places, such as provinces and parishes, since the available place=* values don’t obviously correspond to official designations:
To my knowledge, data consumers only use border_type=* when labeling a place by its type, such as in search results. No renderer uses it to decide how prominently to label a place, and none could be expected to. This is roughly what we’d be communicating to data consumers by making the place=city/town choice hew to official designations: the more sophisticated ones will have to ignore this tagging distinction entirely and rely on something else instead.
This is a problem we’re also grappling with south of the border. Incorporated bedroom communities often have a significant share of a metro area’s population, yet are less prominent in reality than a more isolated place by almost any other measure. I’ve been promoting the idea that, since we already have population=* for the population and border_type=* for the official designation, we could use place=city/town to communicate the difference between a major city and the suburbs that surround it. This approach doesn’t respect official designations, but it doesn’t have to be arbitrary. Statistical concepts like metropolitan areas can inform our decision, even if we don’t map them explicitly as boundaries.
It seems like Albertans have been taking a pragmatic approach for a long time, regardless of any guideline. The other day, I noticed that Gasoline Alley has long been tagged place=suburb, even though neither the province nor the federal government recognizes “suburb” as an official kind of place. I was pretty sure it was intended to represent your average motorist’s notion of Gasoline Alley on both sides of Highway 2. It was upgraded to a suburb in 2014, back when there was no official designation. No one has downgraded it to a hamlet since Red Deer County designated a Gasoline Alley hamlet in 2018. The designated boundary excludes the east side and extends well to the west, but the point is centered over the highway. That’s not to say we couldn’t map a separate Gasoline Alley boundary and tag it as a hamlet, but clearly people are making some affordance for common sense.
Good example.
I would go as far as to say most people are scarcely aware that Gasoline Alley isn’t within the City of Red Deer itself, and presume it’s just an industrial neighbourhood of the City. I honestly didn’t even know Red Deer County created a Gasoline Alley hamlet, and the fact that its boundaries don’t encompass both sides of Highway 2 are silly. The whole area around Highway 2 is in common parlance “Gasoline Alley”. I agree that we could add a hamlet boundary to reflect the official designation, but I would leave the place=suburb node where it is.
Another related point of discussion: what do we do about places that were formerly villages or towns, but voluntarily rescinded those statuses to revert to hamlets? A good example I can think of is Grande Cache. It is still tagged place=townnother mapper added a description “Grande Cache is not incorporated, so it is legally only a hamlet, but it is large enough to be a town.” It was tagged as place=village when it was added to the map in 2008, and changed to place=town in 2014. Officially it was a town from 1983 to 2019, when the town voluntarily unincorporated, downgraded to hamlet, and town council effectively rescinded itself. Even though it’s a hamlet it’s over twice the population of the two towns (Fox Creek and Valleyview) encompassed by the Municipal District of Greenview (the overarching rural municipality). About half the MD’s overall population lives in Grande Cache. The official change was done purely for tax burden reasons; Grande Cache’s population vacillates a lot depending on the financial health of the nearby coal mine and the forestry and oil & gas industries, so rather than be burdened with funding municipal services from a population that was ballooning and shrinking with the whims of the prices of a barrel of oil and dimensional lumber, they rescinded the town and made the tax burden shift to the overarching MD.
1 Like