Talk:Table of United States primary census statistical areas
From Wikipedia, the free encyclopedia
 |
This article is within the scope of the United States WikiProject. This project provides a central approach to United States-related subjects on Wikipedia. Please participate by editing the article, and help us assess and improve articles to good and 1.0 standards. | |
| ??? | This article has not yet received a rating on the Project's quality scale. Please rate the article and then leave a short summary here to explain the ratings and/or to identify the strengths and weaknesses of the article. | |
[edit] neologism
There is no official usage of the term "Primary Census Statistical Area". In fact, virtually all Google search hits are from Wikipedia mirrors. Further, the Office of Management and Budget explicitly says that ranking MSAs and CSAs together is inappropriate as the two entities represent different concepts. --Polaron | Talk 00:36, 10 January 2008 (UTC)
- This article is clearly not an attempt to coin a new term. The expression "primary census statistical area" is always invoked in the lower case and is defined by the following reference:
-
A primary census statistical area is a defined metropolitan or micropolitan region that is not a component of a more populous defined region. In the United States, the 718 primary census statistical areas currently defined by the United States Census Bureau include all 123 Combined Statistical Areas (CSAs) and the 595 Core Based Statistical Areas (CBSAs) that are not a component of a Combined Statistical Area. The Census Bureau defines a Combined Statistical Area as an aggregate of adjacent Core Based Statistical Areas that are linked by commuting ties. The 939 Core Based Statistical Areas currently defined by the Census Bureau include the 363 Metropolitan Statistical Areas (MSAs), which have an urban core population of 50,000 or more, and the 576 Micropolitan Statistical Areas (μSAs), which have an urban core population of 10,000 or more but less than 50,000.
- This article avoids comparing Combined Statistical Areas with Core Based Statistical Areas that may be a component of another Combined Statistical Area. Buaidh (talk) 02:13, 10 January 2008 (UTC)
-
- But CSAs should not be compared to MSAs even if that MSA is not part of a CSA. The two concepts represent different things. Have you ever seen a ranked list published by the federal government that includes both types of entities? If so, please point it out. Also, whether upper case or lower case, the term "primary census statistical area" is still a made-up term and at a minimum should be changed. --Polaron | Talk 05:19, 10 January 2008 (UTC)
- To what??? This article has been here for a year, and you seem to be the first editor to find it objectionable. Does anyone else have a comment? --Buaidh (talk) 15:10, 10 January 2008 (UTC)
- Does anyone besides Polaron wish to delete this article? --Buaidh (talk) 18:29, 11 January 2008 (UTC)
I strongly feel that this table is useful. If we had to choose one or the other, I would keep this table and get rid of the separate tables of MSAs, CSAs, micropolitan areas, etc., which have little significance in isolation.
If the name PCSA (I can't find the reference Buaidh is quoting above, is there a link?) turns out to be unsupported, I would suggest the title Table of United States Census statistical areas.
If including only MSAs etc. that are not part of a CSA is objectionable, an alternative is including all Census statistical areas including CSAs, MSAs, and micropolitan, and adding an additional field listing the enclosing or parent area. The equivalent of the current table could then be viewed by sorting on that column and looking at the rows where the enclosing area field is empty. --JWB (talk) 04:30, 12 January 2008 (UTC)
- That's the main problem. Putting CSAs and CBSAs in the same table together is explicitly mentioned by the OMB to be not appropriate because they represent different concepts. This would be like comparing New England (a region) with New York (a state). Many people already have the misconception that CSAs represent single metropolitan areas when they explicitly are not. We should try and educate people as to the proper use of these statistical areas. How about turning this list into just the CBSAs (which are comparable whether metro- or micro-)? This could then be moved to "Table of Core-based statistical areas", which would be useful as some micropolitan areas are larger than metropolitan areas. --Polaron | Talk 04:56, 12 January 2008 (UTC)
- There is nothing wrong with comparing New England and New York. For some purposes this is more useful as they are of comparable size.
- Could you provide the reference for that OMB statement? Undoubtedly it applies in some circumstances and not in others.
- The CSAs do represent areas that are unified in some important senses - I think the definition is overlapping labor markets. In contrast some of the MSA divisions seem very artificial - for example, Silicon Valley vs. the rest of the Bay Area, or San Bernardino and Riverside counties vs. the rest of the Los Angeles area. --JWB (talk) 05:16, 12 January 2008 (UTC)
- Yes, one could compare a state and a region under certain circumstances - but never when you're ranking states by population. In the same way, we should not rank CSAs and MSAs together here since all we're reporting is population. While the divisions may seem artificial to residents, the delineation rules are applied uniformly throughout the country. CSAs are mainly utilized for calculating unemployment rates, locality pay areas, and also for wholesale commodity distribution, media and labor markets, etc. But that doesn't change the fact that a CSA by definition has multiple central cities. If what this table is trying to do is rank the largest metropolitan areas in the US, then Table of United States Metropolitan Statistical Areas does exactly that. It is of course useful to mention that an MSA is part of a wider region but the wider region should not be used for ranking. See page 10 of this OMB Bulletin to see the statement about not comparing CSAs and MSAs. --Polaron | Talk 05:33, 12 January 2008 (UTC)
- There is nothing wrong with comparing states and regions by population, or states and countries, etc. Of course this is not identical to comparing states alone, but there is no reason why one excludes the other.
- The "multiple center city" definition is arbitrary. SF-Oakland has two center cities, yet is listed as an MSA. Much of the SF-Oakland MSA territory is more integrated with Silicon Valley; the MSA boundaries are simply following county lines. It is impossible to apply rules uniformly across the country when there are huge differences in county size and other factors.
- The Bay Area makes much more sense as a unit, and people compare it as a unit, whether or not it and other comparisons are defined as CSA, MSA, or something else. Are you really trying to forbid anyone from comparing, for example, the Bay Area and the Phoenix area simply because one is a CSA and one is an MSA? This makes no sense at all. --JWB (talk) 06:10, 12 January 2008 (UTC)
- It seems you are trying to restrict discussion to a class of purely formal comparisons that exclude many of the more interesting real-world comparisons. I can't agree with this. --JWB (talk) 06:20, 12 January 2008 (UTC)
- No, the multiple urban core is not arbitrary. San Francisco-Oakland is a single urban area. San Jose is a separate urban area. This is the reason why they are in different MSAs. Yes, because county sizes are not consistent, you do get quirky results sometimes. However, the delineation rules are indeed applied uniformly throughout the country. But be aware that MSAs are only an approximation of the population of a metro area (i.e. the region focused on a single urban area). It usually overestimates one part while underestimates another. Using county subdivisions would be better i my opinion but apparently not all states have commuter worker flow data at that level. Do you think the MSA separation of San Francisco and San Jose was arbitrary? While the there is continuitt between the two physical urban areas, the contact boundary is under the three mile threshold and the commuter flow data show that San Jose is a strong employment attractor. These two reasons satisfy the requirement for splitting urban areas. There is a whole precedure for delineating urban areas and the result of that is the two are separate urban areas. You should read it sometime as it is very interesting. Under certain circumstances, there is no problem comparing the Phoenix metro area with the Bay Area as long as it is clear that the Bay Area consists of six separate metro areas and you are comparing it to a region composed of one. My point is that we must educate people and clearly state that this is not meant as a ranking of metro areas. The MSA definition is the one that is consistent with how other countries define metro areas (a single core area plus adjacent territory closely tied to that core). --Polaron | Talk 14:08, 12 January 2008 (UTC)
- I don't doubt that the boundaries are the result of some algorithm chosen by people at the Census Bureau, rather than directly chosen by people at the Census Bureau. This does not show that the boundaries so defined are the most useful for particular purposes. Also, it's possible to engineer rules to favor chosen outcomes in specific cases.
- Silicon Valley is a powerful employment attractor from the neighboring counties of the SF-Oakland MSA. This is evidence that SF-Oakland and San Jose form an integrated area, not that they are not as you seem to be saying. I'm not sure what you mean by the less than 3 mile contact boundary - the urbanized area of the Peninsula is the same width or greater at the San Mateo-Santa Clara county boundary as it is in the rest of San Mateo County, and appears to be 5 mi or more wide, not even taking into account the boundary with Alameda County on the other side of the bay.
- I don't think there is any problem with readers being unaware that a named area is an MSA, CSA, or something else, as the type is always cited as part of the name.
- It's debatable whether other countries' definitions match the MSA one. First, no two are exactly alike. Second, there are plenty of named areas that are polycentric like the Netherlands Randstad or German Ruhr, and those seem more like the CSAs. You could respond that you have to only compare to the foreign areas that are similar to the MSA definition, but then your reasoning becomes circular. --JWB (talk) 01:13, 13 January 2008 (UTC)
- So you are saying we should ignore the Census Bureau and OMB rules and use our own interpretations? That smacks of original research. If San Francisco and San Jose are so thoroughly integrated, they would have become a single MSA. The reason they are grouped together as a CSA is because they share suburbs, but that doesn't eliminate the fact that there are multiple central urban cores. The urban areas I was referring to are the ones delineated by the U.S. Census Bureau. You can look at maps here. Be aware that there are criteria for hops and jumps. You mentioned the Randstad and Ruhr area. The Randstad is a multiple urban core region and is not normally considered as a single metropolitan area. The Ruhr area is a single urban region, although composed of multiple municipalities. The Ruhr area is part of a wider multicentric region that includes the separate urban areas of Bonn, Cologne, Wuppertal, and Dusseldorf. So, yes, there are similar entries to CSAs elsewhere but those are not usually considered as single metropolitan areas but a series of interacting metropolitan areas.
- How about this: We list the largest CBSAs (regardless of metro or micro) and rank them. This includes CBSAs that are part of a CSA. We insert the CSAs at the appropriate locations in the table but leave them unranked. This way, people can see where the CSA would rank but would be less inclined to use the figure as a basis for ranking. My main concern is people might use CSAs to claim that some city is the Xth largest metropolitan area (in the US, the world, wherever) when the MSA is the most appropriate for comparison.
- No, the multiple urban core is not arbitrary. San Francisco-Oakland is a single urban area. San Jose is a separate urban area. This is the reason why they are in different MSAs. Yes, because county sizes are not consistent, you do get quirky results sometimes. However, the delineation rules are indeed applied uniformly throughout the country. But be aware that MSAs are only an approximation of the population of a metro area (i.e. the region focused on a single urban area). It usually overestimates one part while underestimates another. Using county subdivisions would be better i my opinion but apparently not all states have commuter worker flow data at that level. Do you think the MSA separation of San Francisco and San Jose was arbitrary? While the there is continuitt between the two physical urban areas, the contact boundary is under the three mile threshold and the commuter flow data show that San Jose is a strong employment attractor. These two reasons satisfy the requirement for splitting urban areas. There is a whole precedure for delineating urban areas and the result of that is the two are separate urban areas. You should read it sometime as it is very interesting. Under certain circumstances, there is no problem comparing the Phoenix metro area with the Bay Area as long as it is clear that the Bay Area consists of six separate metro areas and you are comparing it to a region composed of one. My point is that we must educate people and clearly state that this is not meant as a ranking of metro areas. The MSA definition is the one that is consistent with how other countries define metro areas (a single core area plus adjacent territory closely tied to that core). --Polaron | Talk 14:08, 12 January 2008 (UTC)
- Yes, one could compare a state and a region under certain circumstances - but never when you're ranking states by population. In the same way, we should not rank CSAs and MSAs together here since all we're reporting is population. While the divisions may seem artificial to residents, the delineation rules are applied uniformly throughout the country. CSAs are mainly utilized for calculating unemployment rates, locality pay areas, and also for wholesale commodity distribution, media and labor markets, etc. But that doesn't change the fact that a CSA by definition has multiple central cities. If what this table is trying to do is rank the largest metropolitan areas in the US, then Table of United States Metropolitan Statistical Areas does exactly that. It is of course useful to mention that an MSA is part of a wider region but the wider region should not be used for ranking. See page 10 of this OMB Bulletin to see the statement about not comparing CSAs and MSAs. --Polaron | Talk 05:33, 12 January 2008 (UTC)
[edit] Proposed example
>So you are saying we should ignore the Census Bureau and OMB rules and use our own interpretations? That smacks of original research.
- No, I am saying Wikipedia is a resource with data for readers to do their own research, and that we should not try to forbid them from using the government data in real-world ways, for example by insisting on segregating the data in ways that make it much harder to work with all of it.
>If San Francisco and San Jose are so thoroughly integrated, they would have become a single MSA. The reason they are grouped together as a CSA is because they share suburbs, but that doesn't eliminate the fact that there are multiple central urban cores.
- Oakland is a more distinct traditional urban core than San Jose is. Commuting from the East Bay to SF is the older pattern, but commuting from the East Bay (and SF) to Silicon Valley is now extremely large. I would like to look at the algorithm they are actually using (if the Bay Area division is not simply a holdover from earlier years) but at the least I can say the MSA classification does not reflect the way the Bay Area actually works.
>My main concern is people might use CSAs to claim that some city is the Xth largest metropolitan area (in the US, the world, wherever) when the MSA is the most appropriate for comparison.
- I most certainly do claim the Bay Area is the 4th largest urban area in the US (whether you want to call this CSA, megalopolis, megaregion, or anything else) and that the MSA division is not the way people doing real-world comparisons (as opposed to purely formal use of the Census categories) will use the data.
>The urban areas I was referring to are the ones delineated by the U.S. Census Bureau. You can look at maps here. Be aware that there are criteria for hops and jumps.
- I would welcome a pointer to an exact explanation, although this still won't make the MSA organization serve purposes that it actually doesn't.
>You mentioned the Randstad and Ruhr area. The Randstad is a multiple urban core region and is not normally considered as a single metropolitan area. The Ruhr area is a single urban region, although composed of multiple municipalities. The Ruhr area is part of a wider multicentric region that includes the separate urban areas of Bonn, Cologne, Wuppertal, and Dusseldorf. So, yes, there are similar entries to CSAs elsewhere but those are not usually considered as single metropolitan areas but a series of interacting metropolitan areas.
- If it is connected by commuting etc., this is the most important unit and what people are most interested in comparing. Second most often quoted is formal city boundaries, though they have nothing to do with use patterns. I have never seen an analysis treating Santa Clara County as one unit and the rest of the Bay Area as another - it is always the Bay Area as a whole, or an even larger area of Northern California, or individual counties or cities of the Bay Area. The only place I have ever seen this division is in tables of the Census MSA data itself.
>How about this: We list the largest CBSAs (regardless of metro or micro) and rank them. This includes CBSAs that are part of a CSA. We insert the CSAs at the appropriate locations in the table but leave them unranked. This way, people can see where the CSA would rank but would be less inclined to use the figure as a basis for ranking.
- We should facilitate ranking either way. If there is a separate table for MSAs alone, this is an appropriate place for MSA rank; it could also be presented as part of a combined table, though this does not bar the possibility of other rankings of interest. I do not see the point of omitting ranking among all highest-order areas; although it is not very significant for the top few where the results are visible on inspection, it is likely to be of interest for a larger dataset. (We actually seem to agree that strong reader interest exists; we just disagree on whether this established interest should be impeded or facilitated!) --JWB (talk) 17:31, 13 January 2008 (UTC)
Whatever you think doesn't matter as it would be your own interpretation and original research. The MSA definition is how the OMB views immediate commuter flows around a given urban area. CSAs merely show that there is a moderate degree of employment interchange between two adjacent metropolitan areas not that everything is a single metropolitan area. In the end, no one has still shown why two different kinds of entities should be ranked together when the creators of those entities have explicitly said not to do that. Furthermore, you seem transfixed by San Francisco and San Jose. What about Boston and Providence, or New York and New Haven, or Washington and Baltimore? You believe those are single urban cores as well. Seriously, why do you think they weren't merged to a single MSA? The delination rules state that if a central county of one MSA qualifies as an outlying county of another (i.e. if Santa Clara County had a 25% EIM to the SF urban area) then it would become part of the SF MSA. Unless someone can show a population ranking of CSAs and MSAs by the federal government, we should remove CSAs from a ranking of metro areas or include them unranked for comparison.
The main point is that a lot of people (apparently including JWB) have the misconception that CSAs represent single metropolitan areas. They do not or else they would have been a single MSA. Should Wikipedia perpetuate a misconception or should we educate people to utilize these statistical areas in the way they were meant to be? --Polaron | Talk 19:50, 13 January 2008 (UTC)
A CSA is not a Metropolitan Statistical Area, which is a particular Census Bureau definition. This does not say it is not a "metropolitan area" or "urban region" or "conurbation" or "megapolitan area", "megaregion", etc.; it depends on the definitions of each of those terms. To insist these must be equated to the Census Bureau's MSA is a POV, as well as OR if you don't have cites for it - as far as I know, even the Census Bureau does not claim that its particular definitions preempt any definitions of generic terms.
Your viewpoint would also seem to preclude any lists comparing American urban areas to other countries' since the definitions are not exactly the same. Are you advocating deletion of all those articles?
I'm most familiar with the Bay Area-Santa Clara County and LA-Riverside and San Bernardino County examples where it is extremely clear the MSA definitions that follow the arbitrary county lines, mostly reflect the county boundaries rather than any functional separation of the area. As for the East Coast examples, the pairs you cite fall together under some definitions, and for that matter the entire Northeast can and has been considered a megalopolis. All of these viewpoints should be documented; your particular POV does not have the right to exclude them from Wikipedia. --JWB (talk) 20:27, 13 January 2008 (UTC)
- So which figure should one use when one is saying Xth largest metropolitan area? Is it valid to say that Greater New Haven is the biggest metropolitan area in the US? Should one always use the biggest figure that has ever been used that encompasses a given city? Is the BosWash the biggest metropolitan area in the world? Using finer building blocks will always give a more accurate definition but it won't change the fact that the reason why some of these areas are not in the same MSA is that they are in different urban areas.
- To Quote the OMB: "Combined Statistical Areas can be characterized as representing larger regions that reflect broader social and economic interactions, such as wholesaling, commodity distribution, and weekend recreation activities, and are likely to be of considerable interest to regional authorities and the private sector." Comparisons should always be based on a single definition (or as close to one as can reasonably be made). Can you point out other lists on Wikipedia that mix different sources/definitions when ranking entities? If there are, we can fix them to use a single source, otherwise they should be clearly tagged as original research. I'm not sayign we should eliminate the CSA from here but simply ensure that it is not ranked so people don't use it to claim that "but this page says it is the Xth largest metropolitan area". No matter what you say, the CSA represents multiple urban cores while the MSA represents a single urban core. --Polaron | Talk 22:34, 13 January 2008 (UTC)
-
- User:Polaron/Agglomerations mixes CSAs and MSAs, including a dozen or so CSAs, figures for Atlanta as both a CSA and MSA, for Phoenix as a CSA and not a MSA even though Table of United States Combined Statistical Areas lists no Phoenix CSA, and Miami as a MSA only, consistent with Table of United States Combined Statistical Areas having no Miami CSA. It appears that you not only draft mixed lists of the type you are arguing for banning here, but forget that you do so. --JWB (talk) 01:14, 14 January 2008 (UTC)
- Those are from citypopulation.de and are not my figures. I was just trying to determine what definition they were using based on other well-established definitions. As is obviously clear to you, a metropolitan area figure without a definition is meaningless. Any metropolitan area must state what the core area is, what the building blocks are, what the criteria for inclusion are. I am not advocating for using that particular list in an article. Note that Randstad is not there and the Rhine-Ruhr region was subdivided into its component metropolitan areas. —Preceding unsigned comment added by Polaron (talk • contribs) 16:18, 14 January 2008 (UTC)
- And that is a good explanation of why it is necessary to compare various current definitions instead of pretending one exists in isolation. --JWB (talk) 17:43, 14 January 2008 (UTC)
- Comparison of different definitions is fine as long as we don't rank entities based on fundamentally different concepts. --Polaron | Talk 22:05, 14 January 2008 (UTC)
- But you yourself are finding it inevitable. All we can do is label each correctly and prominently.
- Also: granted the CSA grouping criteria are different from the MSA grouping criteria. But what happens when you apply the CSA criteria to an area like Miami? You must get an area identical to the Miami MSA, or else Census Bureau would be listing a Miami CSA with different extent. A list of areas chosen consistently according to the CSA criteria must include the MSAs and μSAs not listed as part of a larger CSA. These must be used together and the caution to not compare CSAs and MSAs must not be addressed to this case. --JWB (talk) 01:48, 15 January 2008 (UTC)
- Comparison of different definitions is fine as long as we don't rank entities based on fundamentally different concepts. --Polaron | Talk 22:05, 14 January 2008 (UTC)
- And that is a good explanation of why it is necessary to compare various current definitions instead of pretending one exists in isolation. --JWB (talk) 17:43, 14 January 2008 (UTC)
- Those are from citypopulation.de and are not my figures. I was just trying to determine what definition they were using based on other well-established definitions. As is obviously clear to you, a metropolitan area figure without a definition is meaningless. Any metropolitan area must state what the core area is, what the building blocks are, what the criteria for inclusion are. I am not advocating for using that particular list in an article. Note that Randstad is not there and the Rhine-Ruhr region was subdivided into its component metropolitan areas. —Preceding unsigned comment added by Polaron (talk • contribs) 16:18, 14 January 2008 (UTC)
- User:Polaron/Agglomerations mixes CSAs and MSAs, including a dozen or so CSAs, figures for Atlanta as both a CSA and MSA, for Phoenix as a CSA and not a MSA even though Table of United States Combined Statistical Areas lists no Phoenix CSA, and Miami as a MSA only, consistent with Table of United States Combined Statistical Areas having no Miami CSA. It appears that you not only draft mixed lists of the type you are arguing for banning here, but forget that you do so. --JWB (talk) 01:14, 14 January 2008 (UTC)
[edit] Length of article
This article is 203 kilobytes long, but I don't think it would be a good idea to split it up. What do you think? --Buaidh (talk) 21:56, 29 February 2008 (UTC)
