The social and political divides along the axis of Rural to Suburban to Urban have been a recurring theme in American (and other) history, which of course continues to this day.
- What is Urban? Suburban? Rural?
- How do we classify the apparent variety and continuum from most urban to most rural?
- Where do you live?
- How does it affect the way you live, what you believe, how you vote?
Apologies for preemptive blog whoring: but similar to last week's regionalism poll, the validity of this poll depends on its being posted as a "Recommended Diary" for at least a full 24 hours in all U.S. time zones, so that it will get an equivalent representative sample of us - morning through night time west coast as well as east coast.
The Urban-Rural divide in America may in its own way be as wide a chasm as Race, Class, Income, Gender, etc. More then the so called Red states and Blue states, some would argue that what there really is is an Urban Archipelago. I believe (i.e. can't find link) that the 50 largest cities, regardless of state, all voted for the Democratic presidential nominee in recent years. Similarly outer suburbs and rural areas of blue states vote Republican. Obviously there are some qualifications such as rural areas with large minority populations yet clearly there are differences.
That said, much of the discussion below focuses on the practical question of how to define urban, suburban and rura in the United States. It is important to remember that both the definition and meaning differs in different countries. I am leaving the discussion of what Urban and Suburban and Rural living means to all of us for the comments section.
Urban:
Urban areas are defined by having some higher level of population and population density. Since 1950s the U.S. Census has used a definition of "Urbanized Areas" as a central core or city and its adjacent, closely settled territory which have a combined total population of 50,000 or more. Supplmenting this, is the suggestion that an "Urban Area" has a density of over 1,000 per square mile.
The concept of Urbanized area as defined by the US Census Bureau are often used as a more accurate gauge of the size of a city, since in different cities and states the lines between city borders and the urbanized area of that city are often not the same. For example, the city of Greenville, South Carolina has a city population under 60,000 but an urbanized area over 300,000, while nearby Winston-Salem, North Carolina has a city population of 180,000, making it appear to be much larger than Greenville, but once the urbanized area is considered, the population is about 270,000 meaning that Greenville is actually a "larger" city for all intents and purposes.
A complete list of the 465 Urbanized areas is available.
Following from this, the term "METROPOLITAN CITY" means:
- a city within a metropolitan area which is the central city of such area, as defined and used by the Office of Management and Budget.
or
- any other city, within a metropolitan area, which has a population of fifty thousand or more.
We will use this definition in our poll.
"Metropolitan Areas" are even larger consisting of a large city and its adjacent zone of influence, or of several neighboring cities or towns and adjoining areas, with one or more large cities serving as its hub or hubs. Unlike an urban area, a metropolitan area includes not only the urban area, but also satellite cities plus intervening rural land that is socio-economically connected to the urban core city, typically by employment ties through commuting, with the urban core city being the primary labor market. The list of 361 Metropolitan Statistical Areas in the United can be seen here.
Suburban:
U.S. government statistics generally use an operational definition for suburban as Metropolitan areas but outside Central city. Suburbs are inhabited districts located either inside a town or city's outer rim or just outside its official limits (the term varies from country to country), or the outer elements of a conurbation. The presence of certain elements (whose definition varies amongst urbanists, but usually refers to some basic services and to the territorial contiguity) identifies a suburb as a peripheral populated area with a certain autonomy, where the density of habitation is usually lower than in an inner city area, though state or municipal house building will often cause departures from that organic gradation. Suburbs have typically grown in areas with an abundance of flat land near a large urban zone, usually with dispersed, less focused or nonexistent city center[1] and with transport technology that allows commuting into more densely populated areas with higher levels of commerce. Suburban communities are not far from urban areas. Families there mostly live in houses in neighborhoods with yards and quiet streets. Many people from the suburbs have to drive to the nearest city to work. Nearby parks and stores support family life. Typically, many post-World War II American suburbs have been characterized by:
- Lower densities than central cities, with single-family homes predominating.
- Zoning patterns that separate residential and commercial development, as well as different intensities and densities of development.
- Subdivisions carved from previously rural land into multiple-home developments built by a single real estate company.
- Shopping malls and strip malls instead of a downtown shopping district.
- Streets lined by off-street car parking and vegetation instead of buildings.
- A road network designed to conform to a hierarchy, including residential streets that curve and often end in culs-de-sac, in place of the grid pattern common to most central cities and pre-World War II suburbs.
- Ready access to freeways or tollways.
- Limited access to public transit.
- The importance of public space reduced in favor of private property.
In addition, suburban areas also encompass pre-World War II smaller towns and cities, as well as other mixed-used activity centers.
Rural:
What is rural? For us, rural areas are those areas...
- that are not in a Metropolitan area at all (Nonmetropolitan area).
- with total population of less than 50,000. No city, town or place within the area has population of 50,000 and total population of area is less than 50,000.
- with large expanses of undeveloped or agricultural land, dotted by small towns, villages, or any other small activity clusters.
- places where there is lots of open countryside. Houses are spaced far apart. There aren't a lot of streets. Traffic is light. There are few stores and businesses.
- Places with populations of less than 2,500
But how to categorize and code this?
Standard Ways to Code:
The best ways to code where a place is along the urban-to-rural continuum requires people to know whether their county is part of a Metroplitan Statistical Area (MSA), as defined by the Office of Management and Budget. Since most folks don't know this information off hand, such a system is not useful for casual self-reporting such as in a dKos poll. Such systems are typically used when one is collecting data such as county name or address or zip code. The the investigator can code afterwards. The system comes from the U.S. Dept of Agriculture but is widely used throughout government and by academics. It is called Rural Urban Continuum codes. They form a classification scheme that distinguishes metropolitan (metro) counties by the population size of their metro area, and nonmetropolitan (nonmetro) counties by degree of urbanization and adjacency to a metro area or areas. The metro and nonmetro categories have been subdivided into three metro and six nonmetro groupings, resulting in a nine-part county codification. The codes allow researchers working with county data to break such data into finer residential groups beyond a simple metro-nonmetro dichotomy, particularly for the analysis of trends in nonmetro areas that may be related to degree of rurality and metro proximity.
You can look up where you live at this link.
There is also a related system that takes the classification-coding scheme down to the zip code level.
What System to Use for Self Coding?
But none of the above is conducive for a daily Kos poll which requires anonymous self coding. The problem with folks deciding whether they are rural or urban on their own is how relative such concepts are for folks. Clearly what is thought of as rural by someone in New York City is very different than a person in a small remote town in Montana. Often, folks do not accurately know the population of the place where they live. There are severe problems associated with how they perceive where they live relative to larger places (e.g., someone can say they live in a place of 10,000 which will turn out to be a sub-community of a large metro area).
So, after some consulting with experts off-dKos, the poll below is what I came up with. It does have the advantage of making only a single division by population size, making your determination easier. It does have the advantage of using the U.S. Census Metro Area and Metro City definition and cut-point of population above or below 50,000. It has the disadvantage of not distinquishing between larger and smaller cities. So a city of 60,000 is lumped together with a population of a million. People who do this work suggest that the real differences are between rural areas and any city, and not so much between a small city and a big city.
And once again, for discussion and comments we ask:
- Urban-Suburban-Rural social and political divides have been a recurring theme in American (and other) history, which of course continue to this day.
- Where do you live?
- How does it affect the way you live, what you believe, how you vote?
For example, many of us think that with Iowa and New Hampshire leading off in the primaries, there is great neglect of urban issues (zoning, mass transit, jobs, housing, schools, poverty, etc).
--------------------------------------------------------------------------
Last Week's Results - Where do You Live by State/Region
The tabl below compares last week's result to that from 2007 and to the U.S. population; make of it what you will:
There is remarkably little change among Daily Kos respondents between 2007 and 2008. This despite overall growth is registered users, primary season, April fools day (OMG did YoU hAte tHaT), Clinton supporters' strike, etc. On the one hand, we appear to still have a relative deficit of southerners. But overall, it appears to be not too far off from being representative geographically in this way the U.S. and certainly not from Democrats. Any characterization of this site as being just northeast, or just the coasts, is clearly false.
* Alas, my attempt to include Puerto Rico and the other non-continental U.S. territories and commonwealths was apparently messed up by the poll button for that option not working during much of the day (as reported in comments and emails). I did not have the "rest of the world" option in 2007.
--------------------------------------------------------------------------
On a Slightly Related Note - I Hate Counties (as a demographic unit):
This is a pet peeve of mine, coming out of my work. Counties are often used as a category when talking about some social or demographic or political discusion or analysis. How many times have you seen a map "by county" in the media? Obviously counties are a commonly used & politically relevant geographic unit. But they have the same sort of "bias" as when States are compared as if they are equivalent units (hey let's give Wyoming and Calfornia equal weight in the Senate).
As of the 2000 Census there were roughly 3,100 Counties in the U.S. But counties are historical-political accidents, and vary enormously by size, both in square miles and in population. Looking at the map of counties below, you can get a sense that there are roughly three kinds of counties: Urban densely populated small in size but relatively large in population; Rural small population and small in size mostly in the east; and Rural small population and large in size mostly in the west.
Continental United States by Counties:
But look at the table of county by population size below, and you will see just how badly counties do not represent people:
It is not just that the population range of County is from just 556 to 9,145,219.
There is County-level bias: Over half of counties have a population of under 25,000 each. Just 34 counties have population of over a million each. Therefore, in any discussion of "by county" there is an over-representation of small population counties. There are more of them, so by chance alone, whatever sort of event or characteristic you are looking at, are more likely to be detected among small population counties.
There is a Person-level bias: If you are counting "by conties", then 7% of the population is over-represented by living in the 50% of counties with the smallest population. 50% of population is under-represented by living in the 4% of counties that are largest in population. 24% of people live in just 1% of all counties (us big city folk).
Counties do not represent communities either. There is the problem of "Internal Heterogeneity". If you just look at county-level data, then you cannot detect differences among sub-populations within a county. This can occur with any county having distinct internal sub-populations, but is more likely in those under-represented counties with large urban population. For example, LA County has within it both Bel Air & South Central. New York County (Manhattan) has Harlem and Park Avenue. A county-wide average, let's say in infant mortality rate, will hide the bimodal distribution of different sub-populations. The high rates in one community lost when averaged with the low rates of another community. And the needs of the high rate community are missed by only seeing the lower average for whole county.
So, county level analysis inherently gives false results. Smaller-area & service-based geographic units are more equivalent and accurate. You can still detect anything going on in small and/or homogeneous and/or rural counties and communities that truly have high rates or are otherwise outliers (using smaller geographic unit is not biased against rural or small counties). But also detect true high rate communities within large population and/or internally heterogeneous counties.
In politics, census tract or precinct level data is often used.
But most people don't know what census tract or precinct they live in.
However, people do know their zip codes!
Zip codes are interesting. They are actually a service route, not technically an area (for proper GIS purposes zip code tabulation areas known as ZCTAs have been created). Simplistically speaking, zip codes are, for the most part, defined by what is a convenient route to cover to deliver the mail, balancing the various issues of geographic size, population size and travel time. In effect a kind of community community. As you can see by the table below they are -- compared to States or Counties -- a nice, smaller, more internally homogenous, and more equivalent unit:
Comparison of Zip Codes, Counties & States:
I like zip code data. In disease, health and health services data, census tract and precicts are usually too small. Zip codes or aggregates such as Primary Care Service Areas (PCSAs) are typically as small as we can get, and as suggested above, far better then county data.
Continental United States by Primary Care Service Areas (PCSAs):
--------------------------------------------------------------------------
The Usual Caveats:
Of course these polls are not "scientific." Since the poll is inherently self selecting and is not a random sample there is always the possibility for selection bias bias including but not limited to responder bias, making the survey results different than truth. These will never be valid random sample. And just size alone does not make it valid since still self selecting and has inherent selection bias. But probably has some validity, though since not a random sample, and don't know enough about total population, sample frame and self selecting factors, cannot even measure how inaccurate it is. Yet probably has some validity... yet not valid and can't even measure how invalid... yet.... Granting all that, there are some indicators suggestive of possible validity in these polls: for example if distriubtion is relatively stable and does not jump around wildly this is not definititive but suggestive (at least that it is not being "freeped". Also, when poll done repeatedly, either same way, or only slightly different way (some advantage to doing slightly differently, if thought out, as test of validity), at different times, lets say different day of week, weekday and weekend, morning vs. night, months or years apart.
Some folks asked why should we do these demographic polls at all, and raised the issue of privacy concerns. As to the why question, the simple answer is "know thyself." Who are we when we spout off and comment? Also, it can be fun. It also helps the discussion of whatever the topic is, in this case, "America by Regions." And yes, maybe it will be reported by other media or be used to market advertising to the site. As to the privacy concerns, Kos has certainly made clear his strong views against "outing" the real identities of anybody. I do not know what access the site administrators have to the data or linkages of usernames to poll responses. Myself, I am just a regular user, and have nothing to do with administration or behind the scenes here. I don't know who has voted or what age goes with whom. All I have is the same bar graph and diary that is publicly visible. Also, there are no cross-tabs between any variables (e.g., prior urban-rural with this week's regional). It is not like a questionnaire with multiple separate questions per single interviewee. I guess the question is a matter of what the site administrators COULD access and link or identify if they were so inclined, whether they WOULD do so; and what protections are there on system to prevent an outsider from doing so? Clearly if there were a serious break of confidentiality/privacy, the Kos community would react very badly. The simple answer is, if you are that concerned, with this or any other issue, then do not participate; don't vote. This is a voluntary poll, within one of many diaries, among the nearly infinite number of webpages you can browse.
--------------------------------------------------------------------------