Template talk:Largest cities of Brazil

Core city

The term core city refers to the central city or the city proper of a metropolitan area (as such, New York City is, obviously and evidently, the core city/city proper of the New York Metropolitan Area). Since this is a table of Brazilian cities, and since the population figures refer to those of cities and not to those of metropolitan areas, it doesn't make sense to have the second column labeled as "core city", instead of simply "city". I suggest that it be changed accordingly. --the Dúnadan 16:59, 5 June 2008 (UTC)

Allegations of incorrect figures from a non-existent census

I am interested in two discrete claims that have just made at WP:AN/I, and only these claims.

First: reverting sourced information and replacing it with incorrect figures - just check the source. I looked at the source. It is dismayingly long for somebody, such as myself, who is ignorant of the geography of Brazil. Please say where the figures may be found within it. (And please do so with no mention of any other editor.)

Secondly: insisting that there is something like a "2008 Census" in Brazil, which is false - the source is clear that those are "estimativas", estimates. I don't speak Portuguese and few people here do. Obviously estimativas is a cognate of "estimates". However, this by itself doesn't prove to me that this is not a census. (For all I know, estimativas could merely be conventionalized modesty, cf internet "Request for Comments".) Please explain. (And again please do so with no mention of any other editor.)

Thank you. -- Hoary (talk) 11:57, 16 May 2009 (UTC)

1. The source is dismayingly long. The best thing to do, since it is a .pdf, is to use Adobe Reader's search function. It will return 4 results, three of which are about small towns (São Paulo do Potengi, São Paulo de Olivença, and São Paulo das Missões), and the remaining is São Paulo, the huge city in question. It clearly shows this figure: 10,990,249 (alternatively, you can look up directly for this figure; just remind that in Portuguese, it is written "10.990.249", with dots, instead of commas).

The source does not mention the figure in the template, and, so, cannot be used to support it. To check this, you can also search for it in the source in the same way. It will give you 0 results.

The figure in the template refers possibly to the Metropolitan Areas instead of the cities themselves. Which makes the template skewed, since for all other cities, except Rio and São Paulo, the figures for the city, not the Metropolitan Areas, are given. This is possibly a POV issue: the text in Demographics of Brazil states that "other Brazilian metropolises (sic) are second tier", which is easier to support with these discrepant figures.

2. Brazilian censuses take place each ten years, in years ending in 0: 1940, 1950, 1960, 1970, 1980, 2000, 2010. The exception is 1991, because budget concerns (and bad policy) postponed it. Or, if we want to go real back in time, 1872.

There is no such thing as a 2008 Census. There are smaller researches, such as the monthly PME (Monthly Research on Employment) and PNAD (Anual Research by Dwelling Samples); but they are not complete as a Census. The PME only researches some metropolitan areas; the PNAD researches a (huge, and representative) sample of the country, and is used to make population estimates, but it is not a Census. Ninguém (talk) 00:21, 19 June 2009 (UTC)

I have found São Paulo (foot of p.71), and the number is indeed 10,990,249. Likewise, Rio de Janeiro (top of p.61) is 6,161,047. I took one other at random; and yes, Belo Horizonte (p.43) is indeed 2,434,642.

They're under the heading municipios, which is surely a cognate of "municipality". Offhand I don't know whether the term corresponds to "metropolitan area" or "city"; I'd guess the latter.

Each number is given to what would be a laughable degree of precision even for a genuine census -- I mean, 10,990,249 at which minute of the day? -- so I don't know how it could be the result of anything less exhaustive than a genuine census. I mean, the numbers may not be the result of a genuine survey, but they also hardly look like intelligent estimates.

Perhaps Opinoso would care to comment. As you have (here) not commented on his motivation, etc, I hope that he does not comment on yours. -- Hoary (talk) 14:06, 16 May 2009 (UTC)

I'm speaking as a disinterested party. The numbers have been changed many times since they were copied from the government statistics. The 2008 numbers, from the cited government source, were added on Sept 11th, 2008. They were initially changed, adding 6 million to Sao Paulo and 5 million to Rio De Janeiro, four days later. Since then, the numbers have varied from -12 to 440 quadrillion, but haven't ever been corrected back until yesterday. The numbers themselves seem to come from a government bureau of statistics and seem to be official estimates, not projections or mathematical estimates. I thought at first that the larger numbers might be inclusive of the mentioned metropolitan areas, but there were no versions that seemed to be a simple sum. If it were up to me, I'd correct the numbers to match the cite and start a pool on how long they last. Celestra (talk) 15:58, 16 May 2009 (UTC)

Well, either correct or incorrect, excessively precise or not excessively precise, those are official numbers by the official Brazilian demographic bureau. If they aren't good enough, I don't think any other numbers are going to be good enough. But regardless, figures should be consistent and comparable: either the figures for metropolitan areas should be used in all cases, or figures for the core city alone should. The way it is, São Paulo and Rio de Janeiro look much more bigger than the others than they really are. In this case, if the quoted .pdf is given as the source, then the figures should match the source. If the .pdf is considered to be substandard, then no figures should be used until a more reliable source is found. Reverting to figures that aren't supported by the source - especially without any explanation of what is being done - does not seem a good idea.

The estimates are either projections from the 2000 Census, taking into account natality, mortality, immigration, and emmigration, or extrapolations from the PNAD, which is a research that covers a huge sample. They probably choose not to round up the figures because they are used to calculate what part of the national budget should be allocated to each city.

The figures seem to have been a favourite delicacy for trolls, who change them to whatever they feel like. I don't know how to address this, but I am pretty certain that driving good faith users away isn't the best idea. Ninguém (talk) 00:21, 19 June 2009 (UTC)

I have reverted the previous edit (inexplicably summarized as "Undid unsourced figures"). The figures now reflect the source that's presented for them. I'm willing to believe that I have made some mistake here, but I want to see a lucid explanation of my mistake right here, on the talk page, before any change is made to the template. I'm inclined to regard any other change to the numbers as vandalism, and am quite prepared to take the usual measures to combat vandalism. -- Hoary (talk) 00:50, 17 May 2009 (UTC)

Recent changes by Rahlgd

Copied from Rahlgd's talk page:

Hi. I'm concerned about the current version of this template. It appears not to have links to any sources and mentions 2006 estimates, but replaced a version with a link to an apparently valid 2008 source. Also, the current version, and the original version you installed in August, have an obvious error for Duque de Caxias, which another editor corrected (or, at least, made more reasonable) and has reversed wikilinks, which another editor corrected. Your most recent change undid those apparent improvements. Can you explain this? As things stand now, it would be better to return to the 2008 sourced version and make whatever detailed improvements you want to make starting with that base. Celestra (talk) 22:54, 8 September 2009 (UTC)

Your edit summary today gave me hope that the problems I pointed out above had been addressed. I am sorry to see that is not the case. Merely changing the date to claim that the values are more current is not helpful since the 2008 estimates do not agree with the table. Please provide a source for your information or back out your changes. I'm copying these messages to the template talk page. Celestra (talk) 20:17, 9 September 2009 (UTC)

If there isn't a source identified by tomorrow, I think we need to revert to the version of the template from early August. Please let me know if you have any thoughts about this one way or the other. Celestra (talk) 20:22, 9 September 2009 (UTC)

Return to single source for the population estimates (and other changes)

Rather than have one table with twenty different sources of information, I've reverted the values back to the ones from the 2008 population estimates from the Brazilian Institute of Geography and Statistics. That document does not have any metropolitan population figures, although several of the city articles claim that as a source (for a variety of different metropolitan totals in some cases). It may be that some people are adding together the individual smaller cities that they "know" are part of the metropolitan area and using that as a metropolitan population, or these numbers may be from some unreferenced source. Since there is a source for the metropolitan populations, I have dropped that column. We can add it back later if a source is found. Finally, I switched the column heading "State" back to "Federative unit", which seems to be the local term for the regions. Celestra (talk) 17:33, 13 September 2009 (UTC)