In the United States, cities often sprawl well beyond their municipal borders. The largest cities in the country often cover several counties, and may even cross state borders. Some cities, like St. Louis, Missouri make up only a small fraction of their encompassing metropolitan populations (a little more than 1/10th it turns out!). These cities are said to be under-bounded; much of the population lives outside the municipal boundary of the primary city. Other cities like Jacksonville, Florida, however, are considered over-bounded; their municipal boundaries are expansive and often include undeveloped areas beyond the urban periphery.
Most cities that fall into the under-bounded category are considered inelastic; they either have weak annexation powers (the ability to incorporate new territory into their jurisdiction) or lack them altogether. Cities that cannot grow by acquiring new territory through annexation (like St. Louis) often struggle financially as they are unable to recapture an ever-shifting tax base.
Here we examine the over- or under-boundedness of the 50 largest U.S. metropolitan areas by calculating the ratio of central city population to metropolitan area population. The central city is generally defined here as the largest city in the metropolitan area. This was easily determined for all but three cases: Virginia Beach-Norfolk, Minneapolis-St. Paul and Dallas-Ft. Worth. Because the two largest cities in these three metros were nearly the same size, we added their populations together. All data was gathered from the latest American Community Survey (2016) conducted by the U.S. Census Bureau.
On average, central city populations accounted for only 27 percent of total metropolitan populations. This varied from a high of 61 for San Antonio, TX to a low of 7 percent for both Riverside-San Bernardino, CA and Miami, FL. The median was similar at 25 percent, and the standard deviation of 14 percent indicates a fairly high degree of dispersion around the mean.
Figure 1 displays the proportion of metropolitan population accounted for by the central city as proportional symbols (i.e., larger circles indicate a larger proportion or residents living in the central city). No definite regional pattern is apparent here; relatively elastic and inelastic cities can be found in all parts of the country.
In Figure 2 (below), city and metropolitan population are graphed on opposing axes, and each city plotted according to their respective populations. New York City stands alone in this graph due to it’s large city and metropolitan populations (8.5 vs. 20 million, respectively). Cities on the right or lower side of the dotted line have lower than average city-to-metropolitan ratios and are probably on the under-bounded side of the spectrum, while those on the left side/above the line are at least less under-bounded than average, with those well above the line likey to be over-bounded. In reality, however, very few cities in the U.S. are truly over-bounded. So let’s take a look at those with particularly high and low ratios.
The two cities with the highest city-to-metro ratios, San Antonio and Jacksonville, have attained their ranking in different ways. San Antonio has remained highly elastic by aggressively annexing surrounding cities and territories, including those that lie beyond its home county of Bexar. Several states in the U.S. do not allow annexation across county borders like this.
The city of Jacksonville, on the other hand, merged with with Duval County in 1968, creating a single consolidated government. Jacksonville is one of only a handful of cities in the U.S. to merge with their host counties. Louisville, also high on our list, is another.
Although aggressive annexation and high elasticity are generally considered positive attributes as they allow a city to maintain their tax base and regional competitiveness, it is possible that too much annexation too quickly can have a detrimental impact. Memphis, for example, is currently considering de-annexing some its territory — territory that may actually cost the city more than it generates in tax revenue. This is often true of low-density, sprawling residential communities that are expensive to maintain. Infrastructure like water and sewer, and services like police and fire, become increasingly costly per capita and per development at low densities.
What about cities at the other end of the spectrum — those with relatively low city-to-metro population ratios? These cities are generally either 1) inelastic with weak or no annexation powers (e.g., St. Louis, Atlanta, Cincinnati) 2) are part of much larger metropolitan areas with substantial populations outside the core county (e.g., Miami, Boston), or both. For so-called “Rust Belt” cities like St. Louis, Detroit, and Cleveland, the lack of strong annexation powers was particularly crippling following the loss of manufacturing jobs and the relocation of the middle class toward the suburban periphery in the late 20th century.
Cities that are severely under-bounded may run into another problem: they often appear much lower on national rankings and appear to be smaller markets than they actually are. Consider Atlanta, for example. It is the 9th most populous metropolitan statistical area in the U.S. with nearly 6 million residents, but the city of Atlanta is only the 38th largest in the country with less than 500k residents. The cities of Mesa, AZ and Fresno, CA are both more populous than Atlanta.
Although Atlanta generally doesn’t have much trouble recruiting businesses, many mid-sized cities of larger metros and combined statistical areas (CSAs) are somewhat handicapped by this city/metropolitan dichotomy. Consider Greensboro, NC. The city has a population of less than 300k, but it is part of a metro with 750k and a CSA with a more competitive 1.6 million. This puts it just ahead of Jacksonville, FL and just behind Virginia Beach-Norfolk, VA. City populations can thus be quite deceiving, and cities with limited elasticity often do a poor job of representing the true population of their urban regions.
See the table below for a complete list of cities/metros and the proportion of metro. population located in the central city.
|City/Metro||% City Pop.|
|San Antonio-New Braunfels, TX||61|
|San Jose-Sunnyvale-Santa Clara, CA||52|
|Louisville/Jefferson County, KY-IN||48|
|Oklahoma City, OK||46|
|Austin-Round Rock, TX||46|
|San Diego-Carlsbad, CA||42|
|New York-Newark-Jersey City, NY-NJ-PA||42|
|Virginia Beach-Norfolk-Newport News, VA-NC||40|
|Milwaukee-Waukesha-West Allis, WI||38|
|Houston-The Woodlands-Sugar Land, TX||34|
|New Orleans-Metairie, LA||31|
|Los Angeles-Long Beach-Anaheim, CA||30|
|Dallas-Fort Worth-Arlington, TX||30|
|Las Vegas-Henderson-Paradise, NV||29|
|Kansas City, MO-KS||23|
|Tampa-St. Petersburg-Clearwater, FL||21|
|Minneapolis-St. Paul-Bloomington, MN-WI||20|
|San Francisco-Oakland-Hayward, CA||19|
|Salt Lake City, UT||16|
|San Juan-Carolina-Caguas, PR||16|
|St. Louis, MO-IL||11|
|Hartford-West Hartford-East Hartford, CT||10|
|Atlanta-Sandy Springs-Roswell, GA||8|
|Riverside-San Bernardino-Ontario, CA||7|
|Miami-Fort Lauderdale-West Palm Beach, FL||7|