Wikipedia:Articles for deletion/List of cities by GDP (PPP) per capita

The following discussion is an archived debate of the proposed deletion of the article below. Please do not modify it. Subsequent comments should be made on the appropriate discussion page (such as the article's talk page or in a deletion review). No further edits should be made to this page.

The result was delete. Sandstein 09:52, 7 January 2022 (UTC)[reply]

List of cities by GDP (PPP) per capita

[Hide this box]New to Articles for deletion (AfD)? Read these primers!

(Find sources: Google (books · news · scholar · free images · WP refs) · FENS · JSTOR · TWL)

This list is a nonsensical conflation of information. It appears at first glance to be a list of all cities over a certain GDP, ordered by GDP, but it is actually cherry-picked from data from individual countries that is not comparable. For example, for this list to be accurate, you would have to believe that there are zero cities in the U.S. with a GDP of over $7,000 but under $20,000 (which is nonsensical). You would similarly have to believe that 100% of the world's cities with a GDP between $7,000 and $15,000 are in either Mexico or Columbia. This list also conflates cities with counties (and, for some reason, the U.S. state of Oklahoma), which are different kinds of geographic entities. It engages in this conflation and misrepresentation of information because it relies on a single source, and draws inferences from that source that can not properly be drawn. The further down the list, the more problematic it becomes for this purpose. This could conceivably be kept if it were limited to some number reasonably susceptible to confirmation by a second source, such as the top 100 or even top 200, but as it stands it is an innacurate synthesis. BD2412 T 05:05, 30 December 2021 (UTC)[reply]

Note: This discussion has been included in the list of Economics-related deletion discussions. BD2412 T 05:05, 30 December 2021 (UTC)[reply]
Hi,
The page could be renamed as "List of OECD metropolitan areas by GDP (PPP) per capita" if you prefer.
If you disagree with the OECD definition of a metropolitan area, then you can complain to them. We don't have to judge whether what the OECD is nonsensical or not. It's a reliable source. A455bcd9 (talk) 05:11, 30 December 2021 (UTC)[reply]
The OECD may well be a reliable source for per capita lists within specific countries, but it is not and therefore can not be used as some kind of master per capita list of the world. It covers a limited number of cities per country, and therefore the list becomes increasingly inaccurate as it descends because the OECD is not attempting to cover what the list purports to cover. That has nothing to do with the definition of a metropolitan area, and everything to do with the synthesis that this list manufactures. BD2412 T 05:20, 30 December 2021 (UTC)[reply]
That's why I suggested renaming it "List of OECD metropolitan areas by GDP (PPP) per capita".
What would be the issue with that? A455bcd9 (talk) 05:25, 30 December 2021 (UTC)[reply]
The list would still be inaccurate. Do you think that 100% of the world's cities with a GDP anywhere between $7,000 and $15,000 are to be found in Mexico and Columbia, and nowhere else? That is what this list says right now, and that's just the tail end. BD2412 T 05:28, 30 December 2021 (UTC)[reply]
Again, I suggested to rename the page as "List of OECD metropolitan areas by GDP (PPP) per capita".
With such a title (limited to OECD, not the world, and to metropolitan areas, not cities) I don't understand why you mention "Do you think that 100% of the world's cities with a GDP anywhere between $7,000 and $15,000 are to be found in Mexico and Columbia, and nowhere else?". A455bcd9 (talk) 05:32, 30 December 2021 (UTC)[reply]
Then the list is just acknowledged as a useless repetition of the single source from which it draws. Furthermore, it would still synthesize the separate lists in a way that the OECD does not. BD2412 T 05:35, 30 December 2021 (UTC)[reply]
"Furthermore, it would still synthesize the separate lists in a way that the OECD does not.": could you please explain your reasoning? I don't understand as anyone clicking on the source can then see the "GDP per capita (USD, constant prices, constant PPP, base year 2015)" in the "Economy" variables. Here's another source using this variable if you want: https://www.oecd-ilibrary.org/sites/reg_cit_glance-2018-35-en/index.html?itemId=/content/component/reg_cit_glance-2018-35-en A455bcd9 (talk) 05:42, 30 December 2021 (UTC)[reply]
The OECD does not present a list of cities by this characteristic that suggests that the bottom 40 cities on the list are found exclusively in two countries. That's what makes this a synthesis (and a clearly incorrect one). It is obvious at the end of the list, but likely to be more problematic higher up in the list, where the missing information is less obvious. Why not just merge the relevant information into the existing (and more substantially sourced) List of cities by GDP? At least then we would have more confidence that the final list was accurate relative to the cities already included in that list. BD2412 T 05:58, 30 December 2021 (UTC)[reply]
The OECD does present a list of cities by this characteristic:
https://stats.oecd.org/Index.aspx?DataSetCode=CITIES# (url: "CITIES")
Then you click on "Economy" then you have "GDP per capita (USD, constant prices, constant PPP, base year 2015)" (go on "Layout" if it's slow to load or hard to display on your device) and this is the same list as this one. With, as you said "the bottom 40 cities on the list are found exclusively in two countries". So again, I don't understand your point here.
"List of cities by GDP" is not a list by GDP per capita. So why would we merge them? A455bcd9 (talk) 06:04, 30 December 2021 (UTC)[reply]
The list in the format presented by the OECD is clearly not the same list, and clearly not comparable. As for merging, it would at least preserve the relevant information in the useful context of an accurate list. BD2412 T 06:23, 30 December 2021 (UTC)[reply]
As far as I remember, I went on that website, exported the list, and added it to Wikipedia. So I don't understand how you can say that "The list in the format presented by the OECD is clearly not the same list, and clearly not comparable". It is the same list in the same format. (the only difference is: they updated the base year from 2010 to 2015 for PPP).
Can you please explain your reasoning? A455bcd9 (talk) 06:28, 30 December 2021 (UTC)[reply]
(Please note that there are 7 pages on that link) A455bcd9 (talk) 06:31, 30 December 2021 (UTC)[reply]
The OECD list to which you have linked does not, as far as I can see, lump all of the cities together. It isolates them by country. If it did lump them, it would give the inaccurate impression (as the article here does) that this is intended to be a list of all cities of the world above a certain GDP in order. Even if the name of this article was changed, it would still give that misleading impression. It is likely useless as a guide to relative per capita GDP beyond the first few dozen items listed, and certainly useless once it gets into the hundreds. BD2412 T 06:34, 30 December 2021 (UTC)[reply]
You just have to click on one year to get the list ordered by GDP per capita. And again, it's a list of OECD metropolitan areas, I don't get why you keep repeating that it's "a list of all cities of the world above a certain GDP in order". In the same way that List of European Union regions by GDP and List of U.S. metropolitan areas by GDP per capita only list cities from, respectively, the EU and the US, that list only contains metropolitan areas from the OECD. You can see that the OECD uses that list in various sources: here and there for instance.
Even though it's limited to the OECD, it's one of the only reliable lists of cities/metropolitan areas by GDP per capita, so it's quite useful. Can you explain why it's useless?
As I wrote in my answer to Elemimele below, here are other sources:
https://pwc.blogs.com/files/global-city-gdp-rankings-2008-2025.pdf (p. 21)
https://www.brookings.edu/research/global-metro-monitor/ ("View data table" then "GDP per capita" or "TABLE 1. Highest and Lowest GDP Per Capita, 300 Largest Metropolitan Economies, 2013" p. 4, PDF)
https://www.brookings.edu/research/global-metro-monitor-2018/ (looks like Brookings publish their Global Metro Monitor every 4 years)
https://www.oxfordeconomics.com/microsites/cities
Therefore, I suggest adding one column per source. Some cities will only be available in one ranking, as in List of cities by GDP. It could also be merged with List of cities by GDP although splitting in two pages would be clearer, and would follow the examples of List of countries by GDP (PPP) per capita vs List of countries by GDP (PPP) and List of U.S. metropolitan areas by GDP per capita vs List of U.S. metropolitan areas by GDP (and probably many other similar pages).
What do you think? A455bcd9 (talk) 11:17, 30 December 2021 (UTC)[reply]
I think the numbers proposed in these different sources don't match, so how are we to determine and present the correct GDP per capita? BD2412 T 18:31, 30 December 2021 (UTC)[reply]
What do you mean by "the numbers don't match"? We don't have to choose the "correct" one, just display data in each column. Similarly in List of countries by GDP (PPP) per capita and List of largest cities there are different figures for each country or city. I don't understand the problem (if there is one). A455bcd9 (talk) 19:41, 31 December 2021 (UTC)[reply]
I mean that "the numbers don't match" in that the different sources list different numbers for given cities, and different orders. Let me approach this from a different angle. At what point does the list cease to be comprehensive, and begin to exclude cities/municipalities because they are over the arbitrary limit for a particular country or region? My sense is that it is somewhere in the top 200, and I think that if this list were to cut off at that point, and multiply sourced for the cities in that range, it would at least be a more accurate and maintainable representation of the title. BD2412 T 21:08, 31 December 2021 (UTC)[reply]
Different numbers and different orders doesn't seem to be a problem to me. We have the same situation in List of countries by GDP (PPP) per capita and List of largest cities.
Brookings only consider the 300 largest urban areas in the world. We could only display the top 200 (or even top 100) as you suggested. Would it be fine for you?
In the case of the OECD (the current version of the article), the list is fully comprehensive: every single OECD metropolitan area is included (but OECD countries only of course). So I think your concern doesn't apply in this case. Am I wrong? A455bcd9 (talk) 11:10, 1 January 2022 (UTC)[reply]
It is not possible or plausible that the U.S. has only 163 metropolitan areas, with none being below $20,000 GDP per capita. There has to be a cut-off point beyond which they are just not being included. BD2412 T 23:29, 1 January 2022 (UTC)[reply]
The US has 384 Metropolitan statistical area. In 2017 (and in 2009 dollars) only 4 of them had a GDP per capital below 20,000$ according to List of U.S. metropolitan areas by GDP per capita (source: bea.gov). This article uses OECD "metropolitan areas" (or "functional urban areas") and has 163 such areas in the US (you can see the map there: https://regions-cities-atlas.oecd.org/subarea/FUA/USA/x/T_T/2018 ). This is explained there: https://www.oecd.org/regional/regional-policy/redefining-urban-9789264174108-en.htm
"The OECD, in cooperation with the EU, has developed a harmonised definition of functional urban areas (FUAs). Being composed of a city and its commuting zone, FUAs encompass the economic and functional extent of cities based on daily people’s movements (OECD, 2012). The definition of FUA aims at providing a functional/economic definition of cities and their area of influence, by maximising international comparability and overcoming the limitation of using purely administrative approaches. At the same time, the concept of FUA, unlike other approaches, ensure a mimimum link to the government level of the city or metropolitan area. The Metropolitan database provides socio-economic and environmental indicators of 691 OECD functional urban areas over 250,000 inhabitants in 36 OECD countries. [...] In the case of European countries they are Local Administrative Units (LAU) according to the terminology adopted by Eurostat. In most cases local units are municipalities. In the case of the United States, census tracts are used for the method, and the final boundaries of the city and the commuting zones are adapted to the county boundaries."
So the OECD definition of a city/metropolitan area/urban area is closer to what the United States Office of Management and Budget calls a combined statistical area (CSA). And there are 172 CSAs in the US.
Coming back to our example, according to the BEA, the MSAs with GDP per capita below 20,000 (2009) USD in 2017 were:
The Villages, FL Metropolitan Statistical Area ($19k)
Homosassa Springs, FL Metropolitan Statistical Area ($19k)
Lake Havasu City-Kingman, AZ Metropolitan Statistical Area ($18k)
Sebring, FL Metropolitan Statistical Area ($18k)
(note that the link point to counties instead of MSAs)
Homosassa Springs, Florida has a population of 14k and The Villages, Florida of 79k people. Given their location and size, the OECD probably included them in what they called the Marion metropolitan area (and what the BEA calls the Ocala, FL MSA). According to the OECD, the Marion metropolitan area had a 2015 GDP per capita of $21k in 2010 dollars, PPP. Given the differences in unit (2009 USD vs 2010 PPP USD) and period (2017 vs 2015) this seems consistent. In any case, the OECD explains that they use data from the BEA to build the ranking so I assume their calculation is correct.
Similarly, the Lake Havasu City-Kingman, AZ Metropolitan Statistical Area is included by the BEA in the Las Vegas–Henderson, NV–AZ CSA. So the OECD probably followed this definition. And as Lake Havasu City has a population of 54,495, its impact on the GDP of Las Vegas is negligible.
Sebring, Florida has a population of 10k people and is so small that it is not included in any US combined statistical area (cf. https://www2.census.gov/geo/maps/metroarea/us_wall/Mar2020/CSA_WallMap_Mar2020.pdf ). Similarly, the OECD didn't include it in its list.
So the results are consistent and it is both possible and plausible that the U.S. has only about 163 cities, with none being below $20,000 GDP per capita, as this is the conclusion of both the OECD and the United States Office of Management and Budget (the latter calling them "combined statistical area").
Could you please explain what the problem is? A455bcd9 (talk) 08:40, 2 January 2022 (UTC)[reply]
For me, delving deeper into this makes it more problematic, not less so. The list, for example, excludes Honolulu, a well-known U.S. state capitol. Why? Because the OECD arbitrarily excludes it from its own list. The same with the less well-known capitol of Boise. More broadly, the OECD only includes 38 countries, so the list arbitrarily excludes the entire continent of Africa, and all of Russia, Argentina, Brazil, China, India, actually all but four countries in Asia and all but three countries in South America. BD2412 T 18:03, 2 January 2022 (UTC)[reply]
According to the Office of Management and Budget, Honolulu is not a combined statistical area. So the OECD doesn't consider it in its list of urban areas either. Whether the OMB and the OECD made a good or bad decision shouldn't be our concern as there are reliable external sources.
Boise is included by the OMB in the Boise City-Mountain Home-Ontario, ID-OR Combined Statistical Area, which the OECD calls "Ada" (cf. the map) as the Ada County, Idaho is the largest county in this urban area.
Therefore, I don't see any issue for these two cities. The OECD is a reliable source—especially for for OECD countries—so I don't think it's our role to fact-check them.
"More broadly, the OECD only includes 38 countries, so the list arbitrarily excludes the entire continent of Africa, and all of Russia, Argentina, Brazil, China, India": that's why we could rename this page "List of OECD metropolitan areas by GDP per capita", to make it clear that this list is limited to cities of the OECD. As far as I know, no one has ever complained that the List of U.S. metropolitan areas by GDP per capita doesn't include Russian, Brazilian, or Indian cities.
So I see two acceptable options here:
Renaming the page "List of OECD metropolitan areas by GDP per capita". Such a list would be similar to List of countries by wealth per adult, or List of U.S. metropolitan areas by GDP per capita, or List of European Union regions by GDP
Keep the current name (or rename as "List of metropolitan areas by GDP per capita") but add other sources such as Brookings' "Global Metro Monitor". OECD numbers would be given for OECD countries only of course. Similarly, the various sources in List of countries by GDP (PPP) per capita and List of countries by life expectancy don't have data for all countries (e.g. Liechtenstein not included in the IMF list, Syria not in the World Bank list, Somalia not in the CIA list, Taiwan not in the WHO list)
What do you think? A455bcd9 (talk) 08:34, 3 January 2022 (UTC)[reply]
The first option is not useful. I don't think it's tenable to have an OECD-only list, because then we're just copying OECD numbers (unless someone wants to do the work of adding and maintaining cities in just the OECD countries that are omitted even from OECD's own lists). The second option is better, but what you would want to do is start from scratch in draft, limit it to a short enough list that it can be easily reviewed and kept up-to-date, and aim to have multiple sources for each claim. BD2412 T 16:33, 4 January 2022 (UTC)[reply]
List of OECD regions by GDP (PPP) per capita is an OECD only list that has been around for almost 10y now. So it seems tenable? (the two could be merged btw)
Also, what do you mean by "adding and maintaining cities in just the OECD countries that are omitted even from OECD's own lists"? If you refer to Turkey: I sent an email about that to the OECD a long time ago and never got an answer...
And yes, I can do that. What about the top 100 as defined by Brookings ( https://www.brookings.edu/research/global-metro-monitor/ )?
Alternative: limit the list to cities with more than 5 million inhabitants according to the UN? This is the same criterion used in List of largest cities. A455bcd9 (talk) 17:10, 4 January 2022 (UTC)[reply]
List of largest cities has 81 cities of 5 million or more. I think we could have a list of 150 or 200, if the list was comprehensive in listing all cities with sufficient GDP's worldwide. BD2412 T 06:26, 5 January 2022 (UTC)[reply]
"all cities with sufficient GDP" or "all cities with sufficient GDP per capita"? A455bcd9 (talk) 07:26, 5 January 2022 (UTC)[reply]
Delete. As amply demonstrated above, this list isn't much more than an out-of-date, partial mirror of another site. Renaming it so that it's obviously a mirror of that site makes it even more pointless, and undermines WP's role as a tertiary source summarising secondary sources, by limiting the list such that it can never be supported by anything but the one source that it mirrors (albeit badly). Elemimele (talk) 10:05, 30 December 2021 (UTC)[reply]
The source offers data up to 2021, so if the issue is that it is out-of-date, I can update it.
Here are other sources:
https://pwc.blogs.com/files/global-city-gdp-rankings-2008-2025.pdf (p. 21)
https://www.brookings.edu/research/global-metro-monitor/ ("TABLE 1. Highest and Lowest GDP Per Capita, 300 Largest Metropolitan Economies, 2013" p. 4)
https://www.brookings.edu/research/global-metro-monitor-2018/ (looks like Brookings publish their Global Metro Monitor every 4 years)
https://www.oxfordeconomics.com/microsites/cities
(Even though we only had the OECD as a source I don't see why it would be a problem, it's like List of European Union regions by GDP or List of U.S. metropolitan areas by GDP per capita)
A455bcd9 (talk) 11:08, 30 December 2021 (UTC)[reply]
Note: This discussion has been included in the list of Lists-related deletion discussions. Spiderone^{(Talk to Spider)} 19:33, 30 December 2021 (UTC)[reply]

Delete per BD2412 and Elemimele. I found this page earlier and could not make sense of it, as it is a WP:SYNTHESIS of the information from the 1 source. Natg 19 (talk) 04:48, 2 January 2022 (UTC)[reply]
I know that the existence of other similar articles cannot be used as a justification to keep a page. However, lists like List of countries by total wealth, List of countries by wealth per adult, EF English Proficiency Index, or List of U.S. metropolitan areas by GDP per capita have been around for a long time (sometimes more than 10 years) and are only a WP:SYNTHESIS of the information from 1 source. So is this a reason good enough to delete a page?
Also, as I explained above, there are other sources that could be added to this page.
So could you please explain why this list should be deleted? A455bcd9 (talk) 08:46, 2 January 2022 (UTC)[reply]

The above discussion is preserved as an archive of the debate. Please do not modify it. Subsequent comments should be made on the appropriate discussion page (such as the article's talk page or in a deletion review). No further edits should be made to this page.