Tag Archives: Charlotte

A Study on Regional Governments


I’m finally finished with the regional governments project that I’ve been working on for something like six weeks and that has kept me from writing in that time.  In the future I may want to use more accurate measures and publish this, so I’m going to practice by presenting this as scientific research.  So here goes.

Introduction

I wanted to study the possibility of creating regional governments in the United States for three reasons.  First, because I agree with The Charter of the New Urbanism when it states that “The metropolitan region is a fundamental economic unit of the contemporary world. Governmental cooperation, public policy, physical planning, and economic strategies must reflect this new reality.” Also, I feel that the lines that divide government designations in America are arbitrary at best and, in many cases, don’t reflect reality on the ground.

It is interesting to compare government designations in Europe and America.  If you look at a map of Europe, you will notice that none of the boundaries are straight lines.  This is because the boundaries do a much better job of reflecting things like topography and real cultural divides.

The US, on the other hand, was drawn up for ease of division by immigrants who considered it to be essentially uninhabited.  Many counties, particularly in the Midwest, are just boxes laid out along the survey lines created by Jefferson, regardless of the topography on the ground.  There is only one state in the US that doesn’t have a straight line for a boundary, and that is Hawaii.  I don’t believe that this is advantageous.  Take my hometown of Pittsburgh, for example.  Pittsburgh is within the state of Pennsylvania, which it shares with Philadelphia.  This is about all the two cities share.  Their economies, populations, ethnic groups, and cultures are very divergent.  When the two cities are thrown into competition, often for State funds, Pittsburgh, with its lower population and generally higher standard of living, often does not get as much as Philadelphia.  According to Pittsburgh Quarterly, “It is often lamented that the Pennsylvania legislature tilts to the east, favoring Philadelphia over Pittsburgh.”  It would be advantageous to Pittsburgh to not be as closely linked to Philadelphia, which it in reality has little relation to.  At the same time, there are areas in Ohio, West Virginia and Maryland that are closer to Pittsburgh than to any major city in their respective states, and would do well to be involved with the politics of that city.  At a lower level, there are people who live just outside of Alleghany County, where Pittsburgh is, so that they don’t have to pay the higher taxes in that county, yet they still use Alleghany County roads and services without paying their fair share.  I believe that, along with state borders, county borders should be amended to reflect the reality on the ground of how central city services are used.

This brings me to my third point: I don’t think, in many parts of the country, that county governments serve a needful purpose in the way that they did in the past.  When the country was made up of many small, independent towns, counties worked to unite them in common purpose.  Now, in our metropolitan world, counties are often used as tax havens or otherwise don’t serve their original purpose.  Rhode Island, Connecticut and Massachusetts have done away with their county governments, streamlining political processes and ironically creating “small government” in some of the most liberal states in the Union.

But how does one define a region?  For an answer, I turned to Christopher Alexander, as I often do, who, in A Pattern Language, encourages us to, “Wherever possible, work toward the evolution of independent regions in the world; each with a population between 2 and 10 million; each with its own natural and geographic boundaries; each with its own economy; each one autonomous and self-governing; each with a seat in a world government, without the intervening power of larger states or countries.”  With that in mind, I began my research.

Methods

I relied heavily on Wikipedia and Google for this research, which is why it isn’t publishable in its current state.  To begin, I set benchmarks for regions.  I wanted to make three maps to visualize regions of different sizes, so I decided that, so as to be in line with Alexander’s requirements, I would have one map with a minimum population of 2 million per region, another map with a minimum of 5 million per region, and a third with a minimum of 8 million.  Then I got a list of cities in the US with a population of over 100,000.

I went through every county in the country and measured the distance from the county seat to the nearest city of over 100,000, marking them on a map with a different color to designate different cities.  I used Google Maps’ walking distance feature because I felt that it would do a better job of reflecting topography than simple as-the-crow-flies measurement, while at the same time being more accurate than the car distances since cars are expected to travel on highways over large distances, which may be faster but not as direct.  Also, I wanted to measure it as if some sort of catastrophe happened and people were unable to use cars, thus being forced to walk.

After finding out which counties were closest to what cities, I counted up the population of the counties that were marked for a given city based on the most recent data on Wikipedia.  Some of this information was as recent as 2009 estimates, while some was as old as the 2000 census.  Hopefully when the new census comes out I can redo this project with better results.  If the population of the area was below the population benchmark that I had set, then the city was removed from the list and the counties were remeasured and marked for the next closest city.  I then mapped them out on large national maps.

Results

This work generated three maps with corresponding lists of cities and the populations of the regions based on these cities.

This first map is based on regions with a minimum population of 2 million, with the following cities anchoring the regions and their given regional populations, from highest to lowest population of the central city:

  1. New York City, NY (10,861,700)
  2. Los Angeles, CA (11.624,092)
  3. Chicago, IL (7,312,584)
  4. Houston, TX (5,807,864)
  5. Phoenix, AZ (6,662,822)
  6. Philadelphia, PA (7,398,857)
  7. San Antonio, TX (3,836,400)
  8. San Diego, CA (3,322,432)
  9. Dallas, TX (3,742,720)
  10. San Jose, CA (3,329,396)
  11. Detroit, MI (2,384,057)
  12. San Francisco, CA (4,318,813)
  13. Jacksonville, FL (3,040,268)
  14. Indianapolis, IN (3,652,091)
  15. Austin, TX (2,079,499)
  16. Columbus, OH (3,736,506)
  17. Fort Worth, TX (3,586,057)
  18. Charlotte, NC (3,878,660)
  19. Memphis, TN (2,502,573)
  20. Boston, MA (6,831,829)
  21. Baltimore, MD (4,243,534)
  22. El Paso, TX (2,874,140)
  23. Seattle, WA (7,070,662) (This includes both Alaska and Hawaii, as will be explained below)
  24. Denver, CO (5,896,137)
  25. Nashville, TN (2,909,035)
  26. Milwaukee, WI (3,184,691)
  27. Washington, DC (3,031,043)
  28. Louisville, KY (2,949,715)
  29. Portland, OR (4,607,152)
  30. Oklahoma City, OK (2,542,568)
  31. Atlanta, GA (6,151,488)
  32. Kansas City, MO (3,814,650)
  33. Fresno, CA (3,032,183)
  34. Sacramento, CA (5,691,903)
  35. Omaha, NE (2,506,874)
  36. Miami, FL (2,785,746)
  37. Cleveland, OH (2,249,989)
  38. Raleigh, NC (2,252,861)
  39. Tulsa, OK (2,843,868)
  40. Minneapolis, MN (4,490,267)
  41. St. Louis, MO (5,069,109)
  42. Tampa, FL (5,049,680)
  43. Santa Ana (Orange County), CA (3,121,251)
  44. New Orleans, LA (2,534,949)
  45. Cincinnati, OH (3,472,024)
  46. Pittsburgh, PA (4,470,907)
  47. Riverside, CA (2,088,322)
  48. Toledo, OH (2,019,458)
  49. St. Paul, MN (2,573,057)
  50. Buffalo, NY (2,782,734)
  51. Greensboro, NC (2,678,241)
  52. Madison, WI (2,070,908)
  53. Orlando, FL (3,625,795)
  54. Birmingham, AL (2,896,134)
  55. Baton Rouge, LA (2,841,516)
  56. Arlington, VA (2,615,764)
  57. Akron, OH (2,307,186)
  58. Montgomery, AL (3,057,149)
  59. Richmond, VA (3,725,124)
  60. Shreveport, LA (2,146,547)
  61. Des Moines, IA (2,092,903)
  62. Augusta, GA (3,286,871)
  63. Grand Rapids, MI (2,311,561)
  64. Little Rock, AR (2,377,037)
  65. Knoxville, TN (3,215,185)
  66. Fort Lauderdale, FL (3,511,282)
  67. Salt Lake City, UT (4,773,812)
  68. San Bernardino, CA (3,454,754)
  69. Fayetteville, NC (2,047,029)
  70. Aurora, IL (3,986,086)
  71. Springfield, MA (3,183,813)
  72. Paterson, NJ (2,285,085)
  73. Syracuse, NY (2,641,398)
  74. Bridgeport, CT (3,876,777)
  75. Warren, MI (2,393,541)
  76. Elizabeth, NJ (4,235,727)
  77. Lansing, MI (2,543,980)
  78. Manchester, NH (3,289,238)
  79. Allentown, PA (2,559,796)

This second map shows regions with a minimum population of 5 million.  They are listed below in the same manner that they were previously.

  1. New York City, NY (15,100,008)
  2. Los Angeles, CA (15,562,860)
  3. Chicago, IL (12,213,121)
  4. Houston, TX (6,568,198)
  5. Phoenix, AZ (8,490,543)
  6. Philadelphia, PA (10,845,050)
  7. San Antonio, TX (5,398,906)
  8. Dallas, TX (5,916,711)
  9. San Jose, CA (9,241,701)
  10. Detroit, MI (8,757,618)
  11. Indianapolis, IN (7,153,419)
  12. Columbus, OH (7,066,082)
  13. Fort Worth, TX (6,422,682)
  14. Charlotte, NC (9,064,119)
  15. Memphis, TN (5,335,220)
  16. Boston, MA (9,275,561)
  17. Seattle, WA (8,755,217)
  18. Denver, CO (10,039,895)
  19. Nashville, TN (5,631,919)
  20. Milwaukee, WI (6,167,922)
  21. Washington, DC (11,269,595)
  22. Portland, OR (5,263,530)
  23. Atlanta, GA (11,943,974)
  24. Kansas City (9,015,985)
  25. Sacramento, CA (6,370,171)
  26. Miami, FL (6,297,028)
  27. Cleveland, OH (6,605,216)
  28. Raleigh, NC (6,911,460)
  29. Minneapolis, MN (8,704,527)
  30. St. Louis, MO (5,438,438)
  31. Tampa, FL (11,235,143)
  32. New Orleans, LA (6,317,469)
  33. Pittsburgh, PA (5,274,967)
  34. Riverside, CA (9,002,191)
  35. Springfield, MA (5,943,610)
  36. Paterson, NJ (6,595,744)

This final map is for regions with a minimum population of 8 million, based on the following cities.

  1. New York City, NY (24,002,264)
  2. Los Angeles, CA (15,562,860)
  3. Chicago, IL (17,521,680)
  4. Houston, TX (14,995,203)
  5. Phoenix, AZ (8,493,518)
  6. Philadelphia, PA (11,376,896)
  7. Dallas, TX (12,594,912)
  8. San Jose, CA (15,669,851)
  9. Detroit, MI (10,047,016)
  10. Indianapolis, IN (14,442,659)
  11. Charlotte, NC (14,787,271)
  12. Memphis, TN (9,796,539)
  13. Boston, MA (12,318,503)
  14. Seattle, WA (13,493,324)
  15. Denver, CO (10,100,944)
  16. Washington, DC (13,875,208)
  17. Atlanta, GA (15,249,097)
  18. Kansas City, MO (10,585,310)
  19. Cleveland, OH (14,161,269)
  20. Minneapolis, MN (9,192,555)
  21. Tampa, FL (17,532,171)
  22. Riverside, CA (9,044,828)

Discussion

There are a number of inferences that can be made from these findings.  The first that I would like to discuss is that, despite using county data, there are still a lot of straight line boundaries.  This is going to be the case as long as counties have boundaries as arbitrary as states.  A more thorough and accurate analysis would include a municipality-by-municipality, rather than county-by-county, analysis, but that would take more time than I am willing to put into this project at this juncture.  The arbitrary straight lines on the map can lead to some unusual results.  For instance, Grand Junction, CO, the county seat of Mesa County, is closer to Salt Lake City and to Denver on the first map, while most of the rest of the counties on the border follow the state line, leaving Mesa County jutting awkwardly into Denver’s region.

Another odd effect is what happens when water transportation is a factor.  Google’s walking directions take regular ferry service into account, so areas such as San Francisco Bay, Puget Sound, Lake Michigan and Massachusetts Bay have many more connections than areas such as Chesapeake Bay.  While in all reality the residents of Northampton County, VA may be more willing to ride a boat to Virginia Beach than to walk to Philadelphia, this isn’t taken into consideration here.  Rivers with infrequent bridges, or at least bridges lacking in pedestrian walkways, also pose a problem.  There are many counties in Arkansas, for instance, that are much closer to Memphis, TN than to Little Rock; however, the lack of bridges and regular ferry service across the Mississippi River made it so that the Google analysis gave many more areas to Little Rock.  Also, Google’s directions from Honolulu to the Mainland included “Kayak across the Pacific Ocean,” and no matter where you wanted the final destination to be, it went through Seattle, thus making Hawaii, as far as this discussion goes, a part of Seattle.

Another issue is the methodology used in selecting which cities would anchor areas.  After having attempted this analysis before with a top-down approach and being unsuccessful, I tried a bottom-up approach, starting with the smallest cities on my list and moving up.  This creates some situations that are somewhat awkward; for instance, Newark, NJ is much more of a population center than either Elizabeth of Paterson, NJ, yet it didn’t make the cut.  Tampa, FL, is another example; it is more likely that Jacksonville and Miami would split the state, rather than Tampa eliminating both of them.  I may in the future consider another top-down approach to see how the results differ.

There is also the fact that this search was limited to cities in the US.  If we were to do a more complete analysis, we would include neighboring countries and, time permitting, the whole world.  There are certainly cities in Alaska, for instance, that are much closer to Vancouver and even Victoria than they are to Seattle.  However, for the purposes of this study, it made sense to limit the scope to the United States.

The last problem with the model is the fact that I set minimum benchmarks.  This worked very well for the first map, which only has two regions exceeding Alexander’s limit of 10 million people, and those not by much.  However, when we get to a minimum of 8 million, nearly all of the regions exceed the limit.  It may be better to next time set a maximum number and split regions in two as they exceed that limit.

These weaknesses being established, there are a few recommendations that I would like to make after doing this research.  First of all, all counties should have one county seat.  There are a number of existing counties that have two seats, and even a few counties that have no seat.  Counties with more than one seat should settle on one and move on, while counties with no seat should either establish one or be dissolved.  Second, if counties are to exist, then all cities should lie within one.  I feel that there is a little bit of leeway in here for state capitals, such as Carson City, NV, which are just following the example of our nation’s capital, but most of the 39 independent cities in Virginia, for example, shouldn’t be independent.  Many of these cities are even the seats of the counties that they are not a part of!  Unless a city has the same boundaries as its county, like Miami and Boston, they should not function independently.  Counties should also be contiguous.  There are a few counties in Louisiana and Kentucky where changing river course or other events have cut certain parts of a county off from the rest of it.  These areas should become part of another, adjacent county.

Also, I will again refer to Christopher Alexander’s A Pattern Language: “Decentralize city governments in a way that gives local control to communities of 5,000 to 10,000 persons. As nearly as possible, use natural geographic and historical boundaries to mark these communities. Give each community the power to initiate, decide, and execute the affairs that concern it closely: land use, housing, maintenance, streets, parks, police, schooling, welfare, neighborhood services.”  While there is a lot in there, Alexander does seem to set 5,000 persons as a baseline for a functional community.  With that being understood, I propose that any counties under 5,000 in population be dissolved.  If this were done, the country would have 292 fewer administrative units to deal with.  The interesting thing is that most of these counties that would go away are not in the sparsely-populated regions of the Rocky Mountains, as I had supposed; they are in the Plains States, where counties were created arbitrarily after Jefferson’s survey and without any sort of requirements for a population to support them.  These counties have no reason to be there, and their citizens would be better off being a part of a real, larger community.

With these suggestions being made, there are still many things that we can learn from these maps.  I personally prefer the first map and think that it could be a good basis for establishing regional governments and possibly eliminating county governments, particularly in the East and in California, where the population is the most dense.  To properly follow the borders of these regions, state borders would also have to be amended.  In this process, States which don’t have significant population centers would be eliminated, including Alaska, Delaware, Hawaii, Idaho, Kansas, Maine, Mississippi, Montana, North Dakota, New Mexico, Nevada, Rhode Island, South Carolina, South Dakota, Vermont, West Virginia and Wyoming.

The second map, with the much fewer and larger regions, might not be as good for establishing regional governments, but may be more useful for realigning state boundaries to better reflect reality.  If this were the plan, then county governments would probably still be needed, but only if they conform to the requirements stated above.  The last map, with the fewest and very largest areas, might not function either as regions or states, but may be one example of how the country might be equitably divided if it were to break up into small countries.  It is interesting to compare this map to others of how the US could potentially break up, as seen here.

Finally, it should be remembered that mere numbers are not what links people to a city or a region.  Few people would ever say that San Jose is the heart of the Bay Area, despite it being considerably bigger than San Francisco.  The only way you would really be able to truly find a dividing line between New York and Boston would be to go door to door through Connecticut and ask people if they are Yankees or Redsox fans.  The only true way to establish a regional identity is through years of tradition and cultural association with an area.  In the words of Lewis Mumford from his epic The City in History, “Contrary to the convictions of census statisticians, it is art, culture, and political purpose, not numbers, that define a city.”

The End of the Cul-de-Sac is Nigh


Lloyd Alter of Treehugger brings us this story on the final days of the cul-de-sac.  He starts by explaining how kids love playing in cul-de-sacs.  Even I will admit that the one time I lived on a cul-de-sac I would spend hours sometimes every day playing hockey in the asphalt circle with the other neighborhood kids.  However, we weren’t allowed to cross the arterial street that our cul-de-sac connected to, because the high speeds of the cars, lack of crosswalks and sidewalks made it very unsafe.  Alter mentions a study from Davis, California, showing that fatal crashes were twice as common in areas built after 1980 than in the old, gridded part of town.  In addition, 59% of trips in the old part of town were made by foot, bike and transit, whereas 14% were made by these modes in the newer parts of town.

Municipalities are beginning to realize the high costs of cul-de-sac development.  Alter quotes a Charlotte, North Carolina study on fire stations in parts of town with high v. low connectivity (i.e. cul-de-sacs):

The least-connected service areas served 5,700 to 7,300 households; the most-connected service areas served 20,800 to 25,900 households. That means there are dramatic differences in the fiscal efficiency of individual fire stations. The stations in least-connected areas cost $586 to $740 per capita annually; the stations in most-connected areas cost $159 to $206 per capita annually.

Virginia has passed a law in the last year that will make cul-de-sacs responsible for their own maintenance, and a number of cities, including Austin, Portland and Charlotte, have passed or are working on ordinances to discourage cul-de-sacs.  I agree with Alter and look forward to the return of the grid in American cities.

Lynx riders trend upward


This article by Steve Harrison shares some interesting information found in a new survey about Charlotte‘s LYNX light rail.  One of the most interesting statistics is that 72% of those surveyed said that they had not used transit before using the light rail service.  This is a great testament to the idea that light rail gets people out of their cars.  LYNX riders had higher average incomes ($65,000) and were better educated (70% having finished college) than their counterparts on express buses ($55,200 and 55%) and regular buses ($31,800 and 25%).  The line passes through and picks up passengers in areas that are both affluent and working class.  “The demographics say it’s a service for everyone,” says Olaf Kinard, a marketing manager for the Charlotte Area Transit System.  LYNX was only expected to serve 9,100 users its first year, but it surpassed that and has topped at about 16,000.  These numbers are very  encouraging for light rail, and I would encourage any transit advocates out there to familiarize themselves with this report and use it in transit discussions.

Will Stimulus Funds Put Transit-Oriented Development Back on the Fast Track?


Randyl Drummer writes a great article on the federal government’s support for TOD.  He says that there has been a gradual move over the years towards TOD, which the stimulus money may help to shove forward.  He sites the examples of transit developments in Seattle, Denver and Charlotte as examples of the future course of TOD.