No Light Rail in Vancouver!
Data Sources and Definitions
Several responses to some of my posts have asked for sources of the data I cite. While it is perfectly appropriate to ask this, the data usually aren’t critical to my main point, which is that planning does more harm than good.
Rather than provide links to every number (which is especially difficult for census data), I would like to list some of my most important sources of data here. I’ll also clarify my use of terms such as “city,” “urbanized area,” etc.
Data Sources
For future reference, many of my data sources are listed in the right column under data:
- References to 1990 or 2000 populations, densities, commuting modes, auto and homeownership rates, and more come from the Census Bureau. Also median incomes and home values for the year before the census (1989, 1999, etc.).
- References to miles of driving come from US DOT’s annual Highway Statistics.
- References to transit ridership and passenger miles usually come from US DOT’s National Transit Database, though data for the most recent year may come from APTA’s Ridership Reports and some cumulative data may come from APTA’s Public Transit Fact Book.
- References to transit’s share of commuter travel come from census data; references to transit’s share of total travel are based on the Highway Statistics and National Transit Data Base.
- More transportation data, such as the mount of travel by train or air, can be found in US DOT’s National Transportation Statistics.
- References to the share of land that is rural, urban, or developed by state are either from census data or USDA’s Natural Resources Inventory. I discuss the relative accuracy of these data in articles here and here.
- The Federal Office of Housing Enterprise Oversight provides accurate data about changes in home prices by metropolitan area.
- In contrast, Coldwell Banker provides interesting information about homes prices in various urban areas relative to each other.
Terminology
As much as possible, I rely on Census Bureau definitions of things like “urban area.” The Census Bureau uses a variety of geographic terms, such as:
- Place means an incorporated city or a Census Bureau-
defined concentration of people. - City means an incorporated city.
- Urban cluster means an incorporated city and the surrounding contiguous suburbs with a total population of 2,500 to 50,000 people. The contiguous development should have a population density of at least 1,000 people per square mile.
- Urbanized area is the same as an urban cluster but with a total population of 50,000 or more. There are more than 400 urbanized areas in the U.S.
- Metropolitan statistical areas are mostly defined using county lines, so may include much rural land. The greater Los Angeles metropolitan statistical area, for example, extends to the Nevada border because the counties are very large.
I try to use urbanized area or urban cluster for most of my data because it is the only sensible comparison of densities, travel modes, etc. Using city would not make sense because some cities (such as Houston and Indianapolis) have annexed most of their suburbs while others (such as Portland and Denver) have not. Using metropolitan statistical area would not make sense because so much of the land in many (but not all) of these areas is rural.
So when I write that Los Angeles is the densest urbanized area in America, don’t send me a comment saying, “How can that be when New York has a population density of 24,000 people per square mile?” The city of New York is denser than the city of Los Angeles, but the New York urbanized area (which includes northeastern New Jersey and southwestern Connecticut) is not as dense as the Los Angeles urbanized area (which includes Pasadena and much of Orange County).
19
Trackback • Posted in Housekeeping
Reprinted from The Antiplanner