Hubness: Redrawing US state lines (quite long)
Grauto
last, I have made enough progress to compose an initial post. If
interest is generated in one or more particular area(s), I'll respond
in greater detail.
The top 48 hubs of the United States, from the many MSAs and cities
whose aerial hubnesses I have computed, are the following:
New York, NY
Los Angeles, CA
Dallas, TX
Chicago, IL
Seattle, WA
Miami, FL
Denver, CO
Atlanta, GA
Billings, MT
Phoenix, AZ
Salt Lake City, UT
Minneapolis, MN
San Francisco, CA
Albuquerque, NM
New Orleans, LA
Boise, ID
Lubbock, TX
Saint Louis, MO
Detroit, MI
Kansas City, MO
Las Vegas, NV
El Paso, TX
Spokane, WA
Houston, TX
Charlotte, NC
Rapid City, SD
McAllen, TX
Cincinnati, OH
Nashville, TN
Fargo, ND
Tampa, FL
Washington, DC
Memphis, TN
Boston, MA
Oklahoma City, OK
San Antonio, TX
Bismarck, ND
Grand Junction, CO
Shreveport, LA
Buffalo, NY
Missoula, MT
Omaha, NE
Sioux Falls, SD
Garden City, KS
Fresno, CA
Pocatello, ID
Springfield, MO
Norfolk, VA
(There might yet be a city missing from this list. In particular,
such a missing city would have to have a hubness higher than 147.31,
the hubness of Norfolk, VA. Let me know if you find a missing city,
and I'll update my data accordingly.)
Now, we can redraw state lines aroudn these 48 hubs, which I will call
_capitals_. I will call these new states _estates_, to distinguish
them from current states.
How do we determine the boundaries of these estates, however? I will
call the most obvious method METHOD 1:
* Given a point X, let H be the capital closest to X. Then H is the
capital of the estate to which X belongs.
I have actually performed this construction, creating a GIF political
map of the United States with these new 48 estates. Not having a
place to put it, I posted it to the Kansas City Highways Yahoo group
(hope you don't mind, TEXAS!), at the following URL:
http://groups.yahoo.com/group/kcroads/
You'll have to join the group; then, the map is filed under the Photos
section. In the meantime, you'll see some other image postings of
mine, most notably some Signmaker fictional gantries and a color-coded
compilation of MoDOT's traffic count data.
If joining Kansas City Highways seems like to much work, e-mail me and
I'll send the GIF as an attachment.
On the map, each crosshair is a capital hub; the house icons are other
high-hubness cities. Some especially notable features of the map:
* The Ozarks gets its own estate, with Springfield as its hub, and the
Joplin and Fayetteville areas as notable cities. This is one of
America's fastest-growing regions, especially relative to surrounding
areas.
* Texas becomes what might be called the Texas Coalition, composed of
six estates: Dallas, Houston, San Antonio, McAllen, Lubbock, and El
Paso.
* California becomes three estates: San Francisco, Fresno, and Los
Angeles.
* Memphis and Cincinnati become, as they ought to be, the centers of
their respective estates.
* New England is defined by a single estate.
* Shreveport, LA gets its own estate. For one who may not understand
why I-69 is being planned along its current alignment, the
hubness-based estates make the benefits clear: McAllen, Houston,
Shreveport, and Memphis are all estate capitals. If one avoids being
distracted by current state lines, I-69 makes more sense.
* The I-29 corridor benefits, with Fargo, Sioux Falls, Omaha, and
Kansas City becoming centers of estates.
* Garden City, KS becomes the center of an estate that, as far as I
can tell, has no MSAs.
One of the best advantages of using hubness to determine estates and
capitals is that the 48 capitals, many of them large metro areas, each
lie within a single estate. This eliminates having to plan across
state lines, a problem in many of America's cities.
However, using METHOD 1 for determining estate boundaries is
problematic, since non-capital hubs can still be split across estate
lines (this occurs in Pittsbugh, PA and Columbia, MO).
Ideally, we would like the estate lines to trace through sparesly
populated areas. I actually have an idea on how to accomplish this!
However, it is somewhat involved; I will denote it METHOD 2.
I'm making some of this up as I go along, so hopefully it will be
coherent/correct. The algorithm is as follows:
* Let X be a point in the United States (latitude-longitude position).
* Define pop(H) to be the population of hub H.
* Define the _order 1 hub_ of X, H_1, to be the closest hub to X.
* Define the _order 2 hub_ of X, H_2, to be the closest hub to H_1
such that pop(H_2) > pop(H_1).
* Define the _order 3 hub_ of X, H_3, to be the closest hub to H_2
such that pop(H_3) > pop(H_2).
* Eventually, H_n, for some n, will be one of the 48 capitals. Then X
is defined to belong to the estate with capital H_n.
(Yes, I'm a math/computer science student!) This algorithm, with a
recursive streak to it, successfully adjusts the boundaries so that
the hinterlands of smaller hubs stay with these hubs, even as parts of
a larger estate. Thus, estate boundaries under this algorithm follow
sparsely populated regions, a very desirable quality.
The above algorithm may seem very difficult to visualize; however, one
can actually perform the algorithm by hand to get a better idea of its
operation. This works as follows: for every non-capital town X,
determine the closest town Y to X such that pop(Y) > pop(X), and draw
a line from X to Y.
If you draw enough lines (it can take quite a few), you'll get an idea
of how my METHOD 2 HUBNESS algorithm above works. Your map will have
a starburst of lines enamating from each capital to some sub-capitals,
with smaller lines enamating from these sub-capitals to
sub-sub-capitals, and so on.
Around each capital, then, a shrublike tree structure forms, with
leaves branching out to the outermost regions of the estate. The
estate boundaries will then be clear as the only remaining whitespace
between estates.
If one starts with the hubness method, then, after determining the
major hubs, constructs this branch-out pattern for each hub, one
immediately sees possible applications of this method, whether for
determining phone/postal districts, or for positioning product
distribution centers.
How does METHOD 2 change the boundaries of the US map I posted to
Kansas City Highways (which was based on METHOD 1)? It's difficult to
say, exactly; it would be a lot of work to get accurate METHOD 2
boundaries, since the boundaries depend on many smaller towns, not
just the capitals themsevles.
However, one can get a general impression of how METHOD 2 boundaries
would be different. The entire Columbia, MO area would become part of
the Saint Louis estate, at the expense of the Kansas City and
Springfield estates. The Omaha estate would probably also grow,
fanning out to the Des Moines hinterlands at the expense of other Des
Moines area estates. This would ultimately be beneficial, uniting
Omaha and Des Moines into a single estate that improves the situation
of Des Moines will still retaining Omaha as the estate capital.
You've finally reached the end of another long-winded hubness post.
This is an amateur analysis; however, the ideas are quite interesting.
It can difficult to think outside the usual perspective of state
lines, and this helps to do just that. I certainly welcome comments,
questions, contentions, and suggestions.
--
John Fischer
math major at Mizzou
Columbia, MO
Pete Jenior
Again, I will comment on the top of the list. It seems as though the
hubness value of a city is reduced if it near other big cities. NYC is able
to overcome this and match its population rank of 1, but Chicago is not and
falls 1 spot to 4. Granted, though, this is what a hub is...
Are you doing this overall now? I though from before that Billings only
made the list if you did things by region?
---
-Pete Jenior - Cincinnati, Ohio
-Freshman Civil Engineering Major
Georgia Tech (downtown Atlanta)
~~~~
"You're on the road
But you've got no destination"
-U2, "Beautiful Day"
"Grauto" <roads...@yahoo.com> wrote in message
news:2ebc118.02031...@posting.google.com...
Andrew Snyder
>
> The top 48 hubs of the United States, from the many MSAs and cities
> whose aerial hubnesses I have computed, are the following:
> Texas Coalition, composed of
> six estates: Dallas, Houston, San Antonio, McAllen, Lubbock, and El
> Paso.
Anyway to post the map without having to join the yahoo group?
Andrew
Bush Lost Newport KY
>* Let X be a point in the United States (latitude-longitude position).
>* Define pop(H) to be the population of hub H.
>* Define the _order 1 hub_ of X, H_1, to be the closest hub to X.
>* Define the _order 2 hub_ of X, H_2, to be the closest hub to H_1
>such that pop(H_2) > pop(H_1).
>* Define the _order 3 hub_ of X, H_3, to be the closest hub to H_2
>such that pop(H_3) > pop(H_2).
This is mighty complicated for what should be a simple matter. To redraw
state boundaries should be based on something like television viewing
areas. There are about 200 DMAs; DMAs that are small in area can be
combined to reduce the number to 50.
We wouldn't think of Dallas as better suited as a hub than Chicago, nor
would we rank Garden City, Kansas, so high on a list of hubs.
Some large DMAs actually take up most or all of a whole state. The Salt
Lake City DMA, for example, usually includes all of Utah.
--
I think. Therefore, I am not a conservative!
------ http://www.todayslastword.org -------
William Lynch
<really big snip>
Just out of curiosity, shouldn't the hub be the largest city
in the new estate? San Francisco flunks.
TEXAS
> map of the United States with these new 48 estates. Not having a
> place to put it, I posted it to the Kansas City Highways Yahoo group
> (hope you don't mind, TEXAS!), at the following URL:
Not at all! Put whatever you like there.
> http://groups.yahoo.com/group/kcroads/
>
> You'll have to join the group; then, the map is filed under the Photos
> section.
I've modified the group settings to where anyone can view the photos.
> * Let X be a point in the United States (latitude-longitude position).
> * Define pop(H) to be the population of hub H.
> * Define the _order 1 hub_ of X, H_1, to be the closest hub to X.
> * Define the _order 2 hub_ of X, H_2, to be the closest hub to H_1
> such that pop(H_2) > pop(H_1).
> * Define the _order 3 hub_ of X, H_3, to be the closest hub to H_2
> such that pop(H_3) > pop(H_2).
>
> * Eventually, H_n, for some n, will be one of the 48 capitals. Then X
> is defined to belong to the estate with capital H_n.
>
> (Yes, I'm a math/computer science student!)
Ouch, my head!
TEXAS
"Andrew Snyder" <jor...@telocity.com> wrote in message
news:3c96e...@nopics.sjc...
> Anyway to post the map without having to join the yahoo group?
The photo section is now available without joining the group.
Robert Coté
roads...@yahoo.com (Grauto) wrote:
> There have been requests for an update to my hubness posts. At long
> last, I have made enough progress to compose an initial post. If
> interest is generated in one or more particular area(s), I'll respond
> in greater detail.
>
> The top 48 hubs of the United States, from the many MSAs and cities
> whose aerial hubnesses I have computed, are the following:
>
> New York, NY
> Los Angeles, CA
> Dallas, TX
> Chicago, IL
> Seattle, WA
> Miami, FL
...
> Fresno, CA
> Pocatello, ID
> Springfield, MO
> Norfolk, VA
>
> (There might yet be a city missing from this list. In particular,
> such a missing city would have to have a hubness higher than 147.31,
> the hubness of Norfolk, VA. Let me know if you find a missing city,
> and I'll update my data accordingly.)
...
>
> * California becomes three estates: San Francisco, Fresno, and Los
> Angeles.
If I've got this right then What happened to San Diego, Santa Barbara,
Bakersfield swallowed up by LA? If San Diego and Bakersfielsd are part
of Los Angeles (250 miles) then Springfield MA and Philly PA are part of
New York.
For that matter Oakland and Sacramento as distinct from SF and Fresno?
As much as I admire the exercise and general idea remember (e)States are
a political animal for pooling common interests and acknowledging
geography. You are correct that Sacramento having anything to do with
Southern California is outmoded and proximity is a factor but it is
-only- one factor. Giving Pocatello Idaho two senate seats for half a
million people and Los Angeles only two seats for 18 million isn't gonna
cut it. You need to -weight- each hub by population (an attractor
function?). Interesting things start to happen then. Providence/Boston/
Hartford/Springfield is an estate and either Pittsfield or Buffalo
become the capitol of Maine/NH/VT and Upper NYS combined.
Brian Reynolds
I wonder why limit to 48 estates.
Perhaps by adjusting the number up or down, the method #1 problem with
Pittsburgh & Columbia would resolve itself.
When you become Prime Minister of the UEA, can I be Minister of
Transportation? There's a bit of tinkering I would like to do with the
Interestate System.
--
Brian Reynolds
Hastings Michigan
Russell Blau
> map of the United States with these new 48 estates. Not having a
> place to put it, I posted it to the Kansas City Highways Yahoo group
> (hope you don't mind, TEXAS!), at the following URL:
>
> http://groups.yahoo.com/group/kcroads/
You might be interested in comparing
http://wireless.fcc.gov/auctions/data/maps/mta.pdf, which divides the
CONUS into 46 areas based mostly on economic criteria. Unlike your
algorithm, these areas follow existing county boundaries.
--
I don't actually have a hotmail account; but I do have one on excite.com
if you really want to get in touch with me.
Joe Morris
> There have been requests for an update to my hubness posts. At long
> last, I have made enough progress to compose an initial post. If
> interest is generated in one or more particular area(s), I'll respond
> in greater detail.
Fascinating. I've been working on similar ideas related to a similar
idea of the City State.... especially as it relates to transportation,
zoning and taxation. Using Estate as the terminology fits right in.
Here in Atlanta, we are getting close to having to force this issue
and have recently developed a few discipline-specific analogs:
GRTA, for example, for transportation is super-regional
http://www.grta.org
I particularly like how some of these Estates worked out on your map:
"Atlanta" works well by tying together a fretful multi-state water issue;
sucking in Chattanooga, Columbus(GA) and Macon really would give them
a voice in things they have little say in now; and beaches on both
the gulf and Atlantic (Woo!)
"New Orleans" makes alot more sense than the current state boundaries
"Dallas" is nice and compact but still brings in Paris(TX) and Waco (I think)
And "Bismark" gets some warmer weather and more of the Missouri River.
Cool ideas
--
Joe Morris, SysAdmin and Not Insane
Atlanta stories http://jolomo.net/
grauto
>other big cities. >NYC is able to overcome this and match its
>population rank of 1, but >Chicago is not and falls 1 spot to 4.
>Granted, though, this is what a hub is...
The cities that benefit most are small, isolated cities such as Garden
City, KS. The cities that benefit least are large cities close to
slightly larger cities, such as Philadelphia, PA. This seems to be the
aspect of hubness that makes people most uncomfortable about it.
>Are you doing this overall now? I though from before that >Billings
>only made the list if you did things by region?
My region labels in the original post may have been misleading. The top
48 hubs have been determined by considering the US as a whole, not
splitting into regions. By splitting into regions, I would have
introduced a personal bias to the results based on my selection of
regions. The hubness list I have produced is based only on the
definition of hubness, and not on any personal preference.
--
Posted at http://www.layover.com/
Trucking jobs, news, features, chat rooms, and more!
grauto
>estate? San Francisco flunks.
I don't use city population; I use metro area population. For San
Francisco, I consider its entire CMSA (including Oakland and San Jose)
to be a single incorporation. With this defintion, San Francisco is the
largest incorporation in its estate.
You bring up an interesting point, though: the largest city in a estate
need _not_ be its capital! This is somewhat rare, but possible. For
instance, Pittsburgh, PA lies in the Buffalo, NY estate, and Hutchinson,
KS lies in the Garden City, KS estate.
grauto
Also, much of central and northern Louisiana belongs to the Shreveport
estate, not the New Orleans estate. If Louisiana is anything like
Illinois, I would imagine that upper Louisiana might enjoy being more
autonomous from New Orleans.
>"Dallas" is nice and compact but still brings in Paris(TX) and Waco
>(I think)
I had to zoom the map out to make it fit in a reasonable size; however,
my original is much more accurate. Dallas does get Paris and Waco, as
well as Temple. Wichita Falls, however, belongs to the Oklahoma City
estate, and Texarkana lies well within the Shreveport estate.
grauto
>>* California becomes three estates: San Francisco, Fresno, and Los
>> Angeles.
>
>If I've got this right then What happened to San Diego, Santa Barbara,
>Bakersfield swallowed up by LA? If San Diego and Bakersfielsd are part
>of Los Angeles (250 miles) then Springfield MA and Philly PA are part
>of New York.
San Diego and Santa Barbara do belong to the Los Angeles estate. I think
Bakersfield also belongs, but just barely; it lies close to the border
of the Fresno estate. If the map were based on METHOD 2, The LA estate
would expand to include Bakersfield's hinterlands.
Philadelphia lies in the New York City estate, but Springfield, MA lies
in the Boston estate. According to my measurements, Hartford, CT lies
just within the Boston estate; however, Albany, NY, lies in the New York
City estate. Syracuse, NY lies in the Buffalo, NY estate.
>For that matter Oakland and Sacramento as distinct from SF >and Fresno?
San Francisco is the first city listed in its CMSA; hence, I take it to
be the capital of an estate. This estate, spanning much of northern
California, contains Oakland and Sacramento. Fresno is the capital of a
separate estate, spanning much of central California.
>As much as I admire the exercise and general idea remember (e)States
>are a political animal for pooling common interests and acknowledging
>geography. You are correct that Sacramento having anything to do with
>Southern California is outmoded and proximity is a factor but it is
>-only- one factor. Giving Pocatello Idaho two senate seats for half a
>million people and Los Angeles only two seats for 18 million isn't
>gonna cut it. You need to -weight- each hub by population (an attractor
>function?). Interesting things start to happen then.
>Providence/Boston/Hartford/Springfield is an estate and either
>Pittsfield or Buffalo become the capitol of Maine/NH/VT and Upper NYS
>combined.
The estates are indeed not ideal for political purposes, due to this
population disparity. It seems to me, however, that they better separate
the US into areas that can better identify culturally. In Missouri, for
instance, the Ozarks region has a character very different from that of
Kansas City or Saint Louis -- and for that matter, Kansas City and Saint
Louis have their share of differences.
I'm particularly interested in how the estate system might be helpful in
highway planning. _Very_ roughly speaking, the longer the border between
two estates, the more likely it is one will find an interstate crossing
this border to connect the two estates. Looking at the estate map,
projects like I-49 and I-69 make much more sense.
These estates are by their very nature oversimplistic; my main intent in
creating them was to try to think of areas of the United States without
being biased by the placement of state lines.
Craig Holl
Obviously Method 1 would have to be tweaked by applying road distance
instead of straight-line distance. One such tweak would be around the
Traverse City, MI area south along Lake Michigan. It is in the Chicago
Estate, but it is closer by auto to Detroit, because of the lake. Also, it
may be preferable to have some boundaries moved to coincide with rivers.
One example I see is on the central western border of Mississippi. The
Shreveport Estate crosses the river, but only slightly. It seems as though
that area should belong to the Memphis and New Orleans areas.
Someone else had the idea of having more or less estates, and see if that
solves any problems, such as the Pittsburg problem.
I like how the I-29 corridor is clearly shown to be important, as I-69.
Another interesting thing to see is how the new capitals diverge around
Oregon. Looking at a US map, you see the interstates doing the same.
Interesting stuff!
--
Craig Holl
Mechanical Engineer
New Berlin, WI
*remove all numbers and caps to reply*
arga...@my-deja.com
> (There might yet be a city missing from this list. In particular,
> such a missing city would have to have a hubness higher than 147.31,
> the hubness of Norfolk, VA. Let me know if you find a missing city,
> and I'll update my data accordingly.)
There is in fact a missing city--my hometown, Wichita, Kansas, whose
hubness is 150. It is the center of a metropolitan statistical area
which includes all of Sedgwick County and has a population of 500,000.
Wichita also supplies the TV feed for nearly all of western Kansas.
The map you have posted on the Yahoo group suggests that Sedgwick
County is split by--and therefore is at the periphery of--three
estates.
> It can difficult to think outside the usual perspective of state
> lines, and this helps to do just that. I certainly welcome comments,
> questions, contentions, and suggestions.
I think the main problem with using a minimum hubness figure to
determine estate capital eligibility is that it leaves in too many
hubs which serve at best an interregional function. It also results
in estates which vary widely in population, which has serious
implications for economic diversity and distribution of rateable
value. The estate for which Garden City is proposed as the capital
would have an almost exclusively agricultural economy in a semi-arid
area where irrigation is required and farming is therefore not
sustainable over the long term. It would have very little urban
rateable value, and a tax base very vulnerable to the business cycle
in agriculture. Southwestern Kansas has periodically contemplated
seceding from the rest of the state, but these initiatives have never
gotten off the ground for precisely these sorts of reasons.
I would suggest a refinement of the estate-creating algorithm along
this general line: the minimum hubness required for a city or M.S.A.
to be an estate capital is that which results in the largest number of
estates such that no estate has a population more than 50% greater
than the average population of each estate.
Scott Nuzum
>can tell, has no MSAs.
That depends. Where did you put Wichita?
It would seem that Garden City would become the smallest capital of the
least populous "estate." That would seem to be one heck of a disadvantage.
S.E.N.
And being from Garden City originally, let me tell you it's got enough
disadvantages.
Scott Nuzum
>state boundaries should be based on something like television viewing
>areas. There are about 200 DMAs; DMAs that are small in area can be
>combined to reduce the number to 50.
>
>We wouldn't think of Dallas as better suited as a hub than Chicago, nor
>would we rank Garden City, Kansas, so high on a list of hubs.
I'm sure you know, but I'll say to the rest of the group that Garden City is
part of the Wichita DMA. Garden City/Dodge City stations basically serve as
repeaters of Wichita stations, although some of them do insert local news
segments (5 to 10 min) into the nightly newscats that oroginate locally.
It's also difficult to imagine Wichita being less important than Garden City
on this list of "hubs". AFAIK, and I could be wrong since I haven't lived
there for 10-plus years, commuter airplanes out of Garden City only flew to
Denver and Wichita.
Of course, three jets, including an Air Canada plane *were* forced to land
there on 9-11. Good thing the Garden City airport used to be a military
base.
S.E.N.
Yep, still a TV geek. =)
Scott Nuzum
TEXAS wrote in message ...
>
>"Grauto" <roads...@yahoo.com> wrote in message
>news:2ebc118.02031...@posting.google.com...
>> I have actually performed this construction, creating a GIF political
>> map of the United States with these new 48 estates. Not having a
>> place to put it, I posted it to the Kansas City Highways Yahoo group
>> (hope you don't mind, TEXAS!), at the following URL:
>
>Not at all! Put whatever you like there.
>
>> http://groups.yahoo.com/group/kcroads/
>>
>> You'll have to join the group; then, the map is filed under the Photos
>> section.
Is this the right URL? I'm getting "cannot find server."
S.E.N.
Of course, this computer is pretty dumb sometimes.
Steve
I got there just fine, must be your computer.
Bush Lost Newport KY
<snu...@terraworld.net> said:
>I'm sure you know, but I'll say to the rest of the group that Garden City is
>part of the Wichita DMA.
Quite frankly, it's silly. Garden City stations should not be lumped in
with Wichita stations when compiling DMAs.
They don't have DMAs for radio, unfortunately, so I've only spent the
past 15 years trying to develop DMA-like areas for radio.
Grauto
> hubness is 150. It is the center of a metropolitan statistical area
> which includes all of Sedgwick County and has a population of 500,000.
> Wichita also supplies the TV feed for nearly all of western Kansas.
I would like to have included Wichita; however, I have calculated its
aerial hubness to be only 129.64, as it is dominated by Tulsa, OK.
> The map you have posted on the Yahoo group suggests that Sedgwick
> County is split by--and therefore is at the periphery of--three
> estates.
METHOD 2 would correct this. The city center of Wichita lies within
the Oklahoma City estate, so METHOD 2 would expand the Oklahoma City
estate to include Wichita and its surrounding areas.
> I think the main problem with using a minimum hubness figure to
> determine estate capital eligibility is that it leaves in too many
> hubs which serve at best an interregional function. It also results
> in estates which vary widely in population, which has serious
> implications for economic diversity and distribution of rateable
> value. The estate for which Garden City is proposed as the capital
> would have an almost exclusively agricultural economy in a semi-arid
> area where irrigation is required and farming is therefore not
> sustainable over the long term. It would have very little urban
> rateable value, and a tax base very vulnerable to the business cycle
> in agriculture. Southwestern Kansas has periodically contemplated
> seceding from the rest of the state, but these initiatives have never
> gotten off the ground for precisely these sorts of reasons.
I mentioned in another post that these estates are indeed
oversimplistic, and are not necessarily politically viable; similarly,
estates are not necessarily economically viable. Although hubness is
loosely population-based, geography alone is the primary
consideration. One could, for instance, double the populations of
both New York and Los Angeles, and the hubness results would remain
unchanged.
It is probably possible to use the mathematical field of topology to
study the structure of the hubness system; then, one might gain a
better understanding of how population and geography interact to yield
the system. The doubling example above, for instance, helps to
illustrate that only relative population -- in particular, population
rank -- is used to determine hubness.
> I would suggest a refinement of the estate-creating algorithm along
> this general line: the minimum hubness required for a city or M.S.A.
> to be an estate capital is that which results in the largest number of
> estates such that no estate has a population more than 50% greater
> than the average population of each estate.
Indeed, my decision to use 48 estates was based only on the fact that
there are currently 48 contiguous states. I haven't tested this
refinement, but it seems that it might not be as helpful in correcting
problems in the hubness system as one might hope. The Billings, MT,
estate, for instance, has a very high hubness, yet is quite sparsely
populated; other estates of its hubness, however, are very densely
populated.
Personally, I would have liked to see estates for such cities as
Wichita, Des Moines, Little Rock, and so on, at the expense of the
most sparsely populated (and possibly not self-supporting) estates;
however, I prefer to stick to the ojective hubness scores, so that my
opinions don't enter into the regions selected.
Michael D. Adams
FCC has had similar thoughts.
The FCC has defined 52 Major Economic Areas in the US. A map is
online at http://www.fcc.gov/oet/info/maps/areas/maps/mea.pdf
The 52 area names are:
Boston
New York City
Buffalo
Philadelphia
Washington
Richmond
Charlotte-Greensboro-Greenville-Raleigh
Atlanta
Jacksonville
Tampa-St. Petersburg-Orlando
Miami
Pittsburgh
Cincinnati-Dayton
Columbus
Cleveland
Detroit
Milwaukee
Chicago
Indianapolis
Minneapolis-St. Paul
Des Moines-Quad Cities
Knoxville
Louisville-Lexington-Evansville
Birmingham
Nashville
Memphis
New Orleans-Baton Rouge
Little Rock
Kansas City
St. Louis
Houston
Dallas-Ft. Worth
Denver
Omaha
Wichita
Tulsa
Oklahoma city
San Antonio
El Paso-Albuquerque
Phoenix
Spokane-Billings
Salt Lake City
San Francisco-Oakland-San Jose
Los Angeles-San Diego
Portland
Seattle
Alaska
Hawaii
Guam & CNMI
Puerto Rico & USVI
American Samoa
Gulf of Mexico
--
___
<*,*> Michael D. Adams | Highway Heaven -- Exit list
[`-'] mda at triskele dot com | index, lists, and more:
-"-"- http://www.triskele.com | http://www.triskele.com/highways
Russell Blau
news:vh7b7a...@linux.triskele.com...
> I don't know if anyone's commented on it here, but it seems that the
> FCC has had similar thoughts.
>
> The FCC has defined 52 Major Economic Areas in the US. A map is
> online at http://www.fcc.gov/oet/info/maps/areas/maps/mea.pdf
>
the Department of Commerce (specifically, the Regional Economic Analysis
Division of the Bureau of Economic Analysis).
As I previously noted, the FCC also uses Rand McNally's Major Trading
Areas (MTAs) as geographic units for some wireless services.
Both of these concepts are based on economic activity, but since
transportation is clearly a factor that affects the cost of goods and
services, one would expect economic regions to have a "hubness"
component.
Joe Morris
> Another interesting thing to see is how the new capitals diverge around
> Oregon. Looking at a US map, you see the interstates doing the same.
Didn't notice that till you pointed it out. It's almost like it's
pushing the capitals away.... an anti-hub. We need another list ;)
> Interesting stuff!
Indeed
Michael Moroney
>I don't know if anyone's commented on it here, but it seems that the
>FCC has had similar thoughts.
>The FCC has defined 52 Major Economic Areas in the US. A map is
>online at http://www.fcc.gov/oet/info/maps/areas/maps/mea.pdf
Not sure how good it is. CA+NV together are split into LA and SF, with LA
having a strange reach across north AZ or S Utah (can't tell which)? And
extremes like North Vermont, Rochester NY and much of northern PA all NYC?
Robert Coté
wrote:
> Thank you for your thoughts; I'll respond below.
>
> >>* California becomes three estates: San Francisco, Fresno, and Los
> >> Angeles.
> >
> >If I've got this right then What happened to San Diego, Santa Barbara,
> >Bakersfield swallowed up by LA? If San Diego and Bakersfielsd are part
> >of Los Angeles (250 miles) then Springfield MA and Philly PA are part
> >of New York.
>
> San Diego and Santa Barbara do belong to the Los Angeles estate. I think
> Bakersfield also belongs, but just barely; it lies close to the border
> of the Fresno estate. If the map were based on METHOD 2, The LA estate
> would expand to include Bakersfield's hinterlands.
>
> Philadelphia lies in the New York City estate, but Springfield, MA lies
> in the Boston estate. According to my measurements, Hartford, CT lies
> just within the Boston estate; however, Albany, NY, lies in the New York
> City estate. Syracuse, NY lies in the Buffalo, NY estate.
Ahhh. Now I see my problem. Well honestly there is no problem just
different methodologies. I'm not interested in the fuzzy details of
edge locations but rather how well your formula reflects actual
conditions on the ground. The short answer is that Primary MA is not
the correct lead agency. SF versus Oakland and Ventura versus Oxnard
being a large and a small example. The second issue is that there
should be a weighting factor for MA size so that very large hub
candidates can acheive hubness of their own despite proximity to a
slightly or even very much larger neighbor. Third there needs to be a
way to pool little hubs to become general hubs above a minimum
population/economic threshold. In this way those places like Wisconson
and the Dakotas can collectively acheive hub status. Finally, you are
mixing two different census tracts CMSA and MSA. Sticking to MSA (MA
now I think) will cure some of the inequalities.
....
>
> The estates are indeed not ideal for political purposes, due to this
> population disparity.
That doesn't necessarially bother me. The disproportionate
representation of the generally common interests of the flyover estates
does. There's too many Pocatellos and no Santa Barbara or San Diego or
Las Vegas or Buffalo/Pittsfield/Springfield/VT. The list goes on.
> It seems to me, however, that they better separate
> the US into areas that can better identify culturally.
At the distinct risk of being labelled an arrogant bicoastalite by most
standards of cultural identitythe rockies to the outskits of Chicago
would be one estate by that measure. Exageration sure but my point
being that I don't think you've made the case that all those little
Midwestern MA centered estates are distinguishable unless you line up
their "Home of the [insert vegetable] Capital of the World" billboards
at the estate border.
> In Missouri, for
> instance, the Ozarks region has a character very different from that of
> Kansas City or Saint Louis -- and for that matter, Kansas City and Saint
> Louis have their share of differences.
>
> I'm particularly interested in how the estate system might be helpful in
> highway planning. _Very_ roughly speaking, the longer the border between
> two estates, the more likely it is one will find an interstate crossing
> this border to connect the two estates. Looking at the estate map,
> projects like I-49 and I-69 make much more sense.
Look how very underserved LA is by that metric. Interstate as
inter-estate commerce has long since been surpassed by intra-estate
mobility and commerce.
> These estates are by their very nature oversimplistic; my main intent in
> creating them was to try to think of areas of the United States without
> being biased by the placement of state lines.
Agreed, interesting and thought provoking and while simplistic there was
no misrepresentation of what you were trying to do. I would however
point out that there are several very strong geographic features that
define many borders and with good reason.
Exile on Market Street
<gtg...@prism.gatech.eduFILTER> wrote:
> Fascinating stuff. Your methods are quite interesting...
>
> Again, I will comment on the top of the list. It seems as though the
> hubness value of a city is reduced if it near other big cities.
And so Philly winds up on the sidelines again...
"So far from God, so close to New York..."
--
Sandy Smith, Univ of Pennsylvania / 215.898.1423 / smi...@pobox.upenn.edu
Managing Editor, _Pennsylvania Current_ cur...@pobox.upenn.edu
Penn Web Team -- Web Editor webm...@isc.upenn.edu
I speak for myself here, not Penn http://pobox.upenn.edu/~smiths/
"Beware of defining as intelligent only those who share your opinions."
--------------------------------------------------------------Ugo Ojetti--
Robert Coté
smi...@pobox.upenn.edu (Exile on Market Street) wrote:
> In article <a76ktc$dfv$1...@news-int.gatech.edu>, "Pete Jenior"
> <gtg...@prism.gatech.eduFILTER> wrote:
>
> > Fascinating stuff. Your methods are quite interesting...
> >
> > Again, I will comment on the top of the list. It seems as though the
> > hubness value of a city is reduced if it near other big cities.
>
> And so Philly winds up on the sidelines again...
>
> "So far from God, so close to New York..."
Why are you repeating yourself?
Personally I appreciate the experiments of Boston, Philly, NYC and DC
while the nation was trying to get it right. Those cities could as well
be renamed in order; Merchantilism, Commerce, Capitalism and Socialism.
Fine tuning and gross failure has led us for better and worse to Los
Angelism. Hey, see any quarter billion abandonded property projects
being proposed for LA? Look above at the list. Boston. CA/T; NYC, WTC;
DC without comment; Philly, the new demolition campaign.
Greg Albright
(the "Crossroads of America") belongs on this list.
Grauto wrote:
> There have been requests for an update to my hubness posts. At long
> last, I have made enough progress to compose an initial post. If
> interest is generated in one or more particular area(s), I'll respond
> in greater detail.
>
> The top 48 hubs of the United States, from the many MSAs and cities
> whose aerial hubnesses I have computed, are the following:
>
> New York, NY
> Los Angeles, CA
> Dallas, TX
> Chicago, IL
> Seattle, WA
> Miami, FL
> Denver, CO
> Atlanta, GA
> Billings, MT
> Phoenix, AZ
>
> Salt Lake City, UT
> Minneapolis, MN
> San Francisco, CA
> Albuquerque, NM
> New Orleans, LA
> Boise, ID
> Lubbock, TX
> Saint Louis, MO
> Detroit, MI
> Kansas City, MO
>
> Las Vegas, NV
> El Paso, TX
> Spokane, WA
> Houston, TX
> Charlotte, NC
> Rapid City, SD
> McAllen, TX
> Cincinnati, OH
> Nashville, TN
> Fargo, ND
>
> Tampa, FL
> Washington, DC
> Memphis, TN
> Boston, MA
> Oklahoma City, OK
> San Antonio, TX
> Bismarck, ND
> Grand Junction, CO
> Shreveport, LA
> Buffalo, NY
>
> Missoula, MT
> Omaha, NE
> Sioux Falls, SD
> Garden City, KS
> Fresno, CA
> Pocatello, ID
> Springfield, MO
> Norfolk, VA
>
> (There might yet be a city missing from this list. In particular,
> such a missing city would have to have a hubness higher than 147.31,
> the hubness of Norfolk, VA. Let me know if you find a missing city,
> and I'll update my data accordingly.)
>
> Now, we can redraw state lines aroudn these 48 hubs, which I will call
> _capitals_. I will call these new states _estates_, to distinguish
> them from current states.
>
> How do we determine the boundaries of these estates, however? I will
> call the most obvious method METHOD 1:
>
> * Given a point X, let H be the capital closest to X. Then H is the
> capital of the estate to which X belongs.
>
> I have actually performed this construction, creating a GIF political
> map of the United States with these new 48 estates. Not having a
> place to put it, I posted it to the Kansas City Highways Yahoo group
> (hope you don't mind, TEXAS!), at the following URL:
>
> http://groups.yahoo.com/group/kcroads/
>
> You'll have to join the group; then, the map is filed under the Photos
> section. In the meantime, you'll see some other image postings of
> mine, most notably some Signmaker fictional gantries and a color-coded
> compilation of MoDOT's traffic count data.
>
> If joining Kansas City Highways seems like to much work, e-mail me and
> I'll send the GIF as an attachment.
>
> On the map, each crosshair is a capital hub; the house icons are other
> high-hubness cities. Some especially notable features of the map:
>
> * The Ozarks gets its own estate, with Springfield as its hub, and the
> Joplin and Fayetteville areas as notable cities. This is one of
> America's fastest-growing regions, especially relative to surrounding
> areas.
>
> * Texas becomes what might be called the Texas Coalition, composed of
> six estates: Dallas, Houston, San Antonio, McAllen, Lubbock, and El
> Paso.
>
> * California becomes three estates: San Francisco, Fresno, and Los
> Angeles.
>
> * Memphis and Cincinnati become, as they ought to be, the centers of
> their respective estates.
>
> * New England is defined by a single estate.
>
> * Shreveport, LA gets its own estate. For one who may not understand
> why I-69 is being planned along its current alignment, the
> hubness-based estates make the benefits clear: McAllen, Houston,
> Shreveport, and Memphis are all estate capitals. If one avoids being
> distracted by current state lines, I-69 makes more sense.
>
> * The I-29 corridor benefits, with Fargo, Sioux Falls, Omaha, and
> Kansas City becoming centers of estates.
>
> * Garden City, KS becomes the center of an estate that, as far as I
> can tell, has no MSAs.
>
> One of the best advantages of using hubness to determine estates and
> capitals is that the 48 capitals, many of them large metro areas, each
> lie within a single estate. This eliminates having to plan across
> state lines, a problem in many of America's cities.
>
> However, using METHOD 1 for determining estate boundaries is
> problematic, since non-capital hubs can still be split across estate
> lines (this occurs in Pittsbugh, PA and Columbia, MO).
>
> Ideally, we would like the estate lines to trace through sparesly
> populated areas. I actually have an idea on how to accomplish this!
> However, it is somewhat involved; I will denote it METHOD 2.
>
> I'm making some of this up as I go along, so hopefully it will be
> coherent/correct. The algorithm is as follows:
>
> * Let X be a point in the United States (latitude-longitude position).
> * Define pop(H) to be the population of hub H.
> * Define the _order 1 hub_ of X, H_1, to be the closest hub to X.
> * Define the _order 2 hub_ of X, H_2, to be the closest hub to H_1
> such that pop(H_2) > pop(H_1).
> * Define the _order 3 hub_ of X, H_3, to be the closest hub to H_2
> such that pop(H_3) > pop(H_2).
>
> * Eventually, H_n, for some n, will be one of the 48 capitals. Then X
> is defined to belong to the estate with capital H_n.
>
> (Yes, I'm a math/computer science student!) This algorithm, with a
> recursive streak to it, successfully adjusts the boundaries so that
> the hinterlands of smaller hubs stay with these hubs, even as parts of
> a larger estate. Thus, estate boundaries under this algorithm follow
> sparsely populated regions, a very desirable quality.
>
> The above algorithm may seem very difficult to visualize; however, one
> can actually perform the algorithm by hand to get a better idea of its
> operation. This works as follows: for every non-capital town X,
> determine the closest town Y to X such that pop(Y) > pop(X), and draw
> a line from X to Y.
>
> If you draw enough lines (it can take quite a few), you'll get an idea
> of how my METHOD 2 HUBNESS algorithm above works. Your map will have
> a starburst of lines enamating from each capital to some sub-capitals,
> with smaller lines enamating from these sub-capitals to
> sub-sub-capitals, and so on.
>
> Around each capital, then, a shrublike tree structure forms, with
> leaves branching out to the outermost regions of the estate. The
> estate boundaries will then be clear as the only remaining whitespace
> between estates.
>
> If one starts with the hubness method, then, after determining the
> major hubs, constructs this branch-out pattern for each hub, one
> immediately sees possible applications of this method, whether for
> determining phone/postal districts, or for positioning product
> distribution centers.
>
> How does METHOD 2 change the boundaries of the US map I posted to
> Kansas City Highways (which was based on METHOD 1)? It's difficult to
> say, exactly; it would be a lot of work to get accurate METHOD 2
> boundaries, since the boundaries depend on many smaller towns, not
> just the capitals themsevles.
>
> However, one can get a general impression of how METHOD 2 boundaries
> would be different. The entire Columbia, MO area would become part of
> the Saint Louis estate, at the expense of the Kansas City and
> Springfield estates. The Omaha estate would probably also grow,
> fanning out to the Des Moines hinterlands at the expense of other Des
> Moines area estates. This would ultimately be beneficial, uniting
> Omaha and Des Moines into a single estate that improves the situation
> of Des Moines will still retaining Omaha as the estate capital.
>
> You've finally reached the end of another long-winded hubness post.
> This is an amateur analysis; however, the ideas are quite interesting.
Grauto
> (the "Crossroads of America") belongs on this list.
I have defined the hubness of a city to be the distance to the closest
larger city. My previous posts give this definition in more detail.
Indianapolis scores well in its region; however, it does not make the
national top 48 hubness list because it is overshadowed by Cincinnati,
OH. Indianapolis lies in the Cincinnati estate.
Under METHOD II, Indianapolis would gain a more prominent position in
the Cincinnati estate, effecting something of a 'twin city' estate not
dissimilar to the Oklahoma City-Wichita or Omaha-Des Moines METHOD II
estates.
William Lynch
OK, I'm still confused here. Are we talking about city size, size
of the MSA, or size of the CMSA. Under two of the three, San
Jose is larger than anything around except Los Angeles, but you
keep sweeping it under the rug as 'just one of those stupid western
things'. There are some serious flaws with this entire process.
Grauto
> of the MSA, or size of the CMSA. Under two of the three, San
> Jose is larger than anything around except Los Angeles, but you
> keep sweeping it under the rug as 'just one of those stupid western
> things'. There are some serious flaws with this entire process.
The process is not necessarily meant to effect any kind of serious
change; for my purposes, it is just a visualization tool.
First, I consider each Census Bureau CMSA to be a single
incorporation, with the first listed city as the name of the
incorporation. Then, I consider each remaining MSA to be a single
incporporation, with the first listed city as the name of the
incorporation. Finally, each non-MSA city/CDP earns a position as a
separate incorporation.
To construct the hubness list, I used a Census Bureau document listing
the 280 CMSAs/MSAs in the United States. The document will be
unsightly here, since the margins are too wide, but I'll paste in the
header and top fifteen cities for reference.
*****
Census 2000 PHC-T-3. Ranking Tables for Metropolitan Areas: 1990 and
2000
Table 3: Metropolitan Areas Ranked by Population: 2000
Note: 1990 Census population counts are as published in 1990 census
reports and do not include changes published subsequently due to
boundary or other changes. Metropolitan Areas are as defined on June
30, 1999 by the Office of Management and Budget. Eight new
metropolitan areas were announced between 1993 and 1999:
Auburn-Opelika, AL MSA; Corvallis, OR MSA; Flagstaff, AZ-UT MSA; Grand
Junction, CO MSA; Hattiesburg, MS MSA; Jonesboro, AR MSA; Missoula, MT
MSA: Pocatello, ID MSA.
Source: U.S. Census Bureau
Internet Release date: April 2, 2001
For information on confidentiality protection, sampling error,
nonsampling error, and definitions, see
http://factfinder.census.gov/home/en/datanotes/expplu.html.
-----------------------------------------------------------------------------------------------------------------------------------------
MSA/ | |
| Census Population | Change, 1990 to 2000
CMSA | |
|-------------------------------|-----------------------
Code | Rank | Area Name
| April 1, 2000 | April 1, 1990 | Number | Percent
-----------------------------------------------------------------------------------------------------------------------------------------
5602 1 New York--Northern New Jersey--Long Island,
NY--NJ--CT--PA CMSA 21,199,865 19,549,649 1,650,216
8.4%
4472 2 Los Angeles--Riverside--Orange County, CA CMSA
16,373,645 14,531,529 1,842,116 12.7%
1602 3 Chicago--Gary--Kenosha, IL--IN--WI CMSA
9,157,540 8,239,820 917,720 11.1%
8872 4 Washington--Baltimore, DC--MD--VA--WV CMSA
7,608,070 6,727,050 881,020 13.1%
7362 5 San Francisco--Oakland--San Jose, CA CMSA
7,039,362 6,253,311 786,051 12.6%
6162 6 Philadelphia--Wilmington--Atlantic City, PA--NJ--DE--MD
CMSA 6,188,463 5,892,937 295,526 5.0%
1122 7 Boston--Worcester--Lawrence, MA--NH--ME--CT CMSA
5,819,100 5,455,403 363,697 6.7%
2162 8 Detroit--Ann Arbor--Flint, MI CMSA
5,456,428 5,187,171 269,257 5.2%
1922 9 Dallas--Fort Worth, TX CMSA
5,221,801 4,037,282 1,184,519 29.3%
3362 10 Houston--Galveston--Brazoria, TX CMSA
4,669,571 3,731,131 938,440 25.2%
0520 11 Atlanta, GA MSA
4,112,198 2,959,950 1,152,248 38.9%
4992 12 Miami--Fort Lauderdale, FL CMSA
3,876,380 3,192,582 683,798 21.4%
7602 13 Seattle--Tacoma--Bremerton, WA CMSA
3,554,760 2,970,328 584,432 19.7%
6200 14 Phoenix--Mesa, AZ MSA
3,251,876 2,238,480 1,013,396 45.3%
5120 15 Minneapolis--St. Paul, MN--WI MSA
2,968,806 2,538,834 429,972 16.9%
Pete Jenior
> > OK, I'm still confused here. Are we talking about city size, size
> > of the MSA, or size of the CMSA. Under two of the three, San
> > Jose is larger than anything around except Los Angeles, but you
> > keep sweeping it under the rug as 'just one of those stupid western
> > things'. There are some serious flaws with this entire process.
>
> The process is not necessarily meant to effect any kind of serious
> change; for my purposes, it is just a visualization tool.
>
> First, I consider each Census Bureau CMSA to be a single
> incorporation, with the first listed city as the name of the
> incorporation. Then, I consider each remaining MSA to be a single
> incporporation, with the first listed city as the name of the
> incorporation. Finally, each non-MSA city/CDP earns a position as a
> separate incorporation.
>
> To construct the hubness list, I used a Census Bureau document listing
> the 280 CMSAs/MSAs in the United States. The document will be
> unsightly here, since the margins are too wide, but I'll paste in the
> header and top fifteen cities for reference.
>
> *****
>
> Census 2000 PHC-T-3. Ranking Tables for Metropolitan Areas: 1990 and
> 2000
It looks like the 2000 figures are finally out! I will have to check this
out...
[snip]
---
-Pete Jenior - Cincinnati, Ohio
-Freshman Civil Engineering Major
Georgia Tech (downtown Atlanta)
~~~~
"You're on the road
But you've got no destination"
-U2, "Beautiful Day"
Scott Nuzum
Bush Lost Newport KY wrote in message <3c98086a.31455683@localhost>...
>On Tue, 19 Mar 2002 18:21:47 -0600, "Scott Nuzum"
><snu...@terraworld.net> said:
>
>>I'm sure you know, but I'll say to the rest of the group that Garden City
is
>>part of the Wichita DMA.
>
>Quite frankly, it's silly. Garden City stations should not be lumped in
>with Wichita stations when compiling DMAs.
>
>They don't have DMAs for radio, unfortunately, so I've only spent the
>past 15 years trying to develop DMA-like areas for radio.
I'd agree if Garden City's stations were airing different programming. But
all the other stations in Wichita-Hutchinson Plus are just glorified
repeaters, airing the exact same signal as the Wichita stations. This is why
it's all one large DMA.
Back when TV was coming out to western Kansas, I think most of the stations
were put on the air by Wichita stations. Only Hays/Goodland stations KAYS
and KLOE (as they were then called) were the exceptions (they had one owner
in Hays). The owner of this station found it was less expensive to air some
programming by repeating the signal from the Wichita CBS station, then
called KTVH, than it was to get it on his own. Eventually, the Hays/Goodland
duo were bought by KTVH (now KWCH) and started running KTVH's signal all day
long, not just in certain hours (Up until then KAYS/KLOE were CBS/ABC
stations; once they became KTVH's property, they dropped the ABC
affiliation).
The current set up hurts the Wichita Fox, UPN and WB stations because none
of the three currently serve the entire DMA. The Fox station only has one
other full-power station (in Salina; which is that city's only full-power TV
station) and three translators (Great Bend/Hays, Dodge City and Garden
City). UPN and WB don't even have that, although they do have cable
coverage.
S.E.N.
Normally, Wichita-Hutchinson would be four different DMAs.
Doug Weasner
has 5 outbound intertstates: (71 north & south 74 west, and 75 N&S). Indy
has seven: (65 N&S, 69 N, 70 E&W, and 74 E&W) , soon to be 8 when I-69 gets
built. As a matter of fact, there's nowhere you can get from Cincy (Dayton,
Columbus, Louisville, and Indy) that you can't get to from Indy just as
easily. Lexington is a slight exception, but here comes my rebuttal: Indy
has Ft. Wayne, Lansing, St. Louis, Chicago, Peoria and soon to be Evansville
to boot. That's a pretty compelling case for putting Indy ahead of Cincy,
IMO.
--
Doug Weasner
Greenwood IN
I want I-69 in MY backyard
"Grauto" <roads...@yahoo.com> wrote in message
news:2ebc118.02032...@posting.google.com...
Pete Jenior
"Doug Weasner" <redball...@hotmail.com> wrote in message
news:u4wn8.992$Mx3.1...@news1.iquest.net...
> Indianapolis ownz Cincinnati as far as transportation signifcance. Cincy
> has 5 outbound intertstates: (71 north & south 74 west, and 75 N&S). Indy
> has seven: (65 N&S, 69 N, 70 E&W, and 74 E&W) , soon to be 8 when I-69
gets
> built. As a matter of fact, there's nowhere you can get from Cincy
(Dayton,
> Columbus, Louisville, and Indy) that you can't get to from Indy just as
> easily. Lexington is a slight exception, but here comes my rebuttal: Indy
> has Ft. Wayne, Lansing, St. Louis, Chicago, Peoria and soon to be
Evansville
> to boot. That's a pretty compelling case for putting Indy ahead of Cincy,
> IMO.
>
Indy may seem like more of a hub than Cincinnati because it is the
undisputed main city in Indiana, but I think this is what John is trying to
get us to think beyond. Granted, there are enough sizable cities in the
midwest that it is hard to pick a hub.
---
-Pete Jenior - Cincinnati, Ohio
-Freshman Civil Engineering Major