How did the framers of the original Interstates decide what cities to include?

[bandit957]

When the Interstate highway system was first laid out, how did they decide what cities would be included, or what cities would be linked with each other?

Memphis and Birmingham were both big cities then, but they weren't linked to each other yet. Des Moines was linked to Kansas City but not St. Louis. Why no link from Columbus to Toledo? I'm sure much smaller cities were included to be linked to each other. Arizona had a lot of Interstates for being such a sparsely populated state back then.

Did the cities mentioned above just not have historic links with each other? Was it because other states had more powerful congresscritters?

[TheHighwayMan3561]

It seemed a lot of the planning was based on existing US routes, and since there was never a direct route from Des Moines to St. Louis that's probably a part of it.

[froggie]

If one reads into the 1944 Interregional Highways report, whose recommendation of ~34K miles served as the backbone of today's Interstates, one will find that it was a very data-driven investigation, trying to maximize the service of the system while minimizing the length (and thus cost).  There were a number of factors involved...I'll touch upon two of them here:

- The 34K system was the smallest system that served all cities of 300K or higher population (as of the 1940 census).  It also served all but three of the 100-300K cities, which was better than some of the larger systems studied.

- The recommended system had the highest forecasted average daily traffic per mile of the systems studied.  This was their key measure of efficiency of the system.  If you can serve more traffic with a smaller system, it stands a better chance of getting funded.

[Rothman]

Data in 1944 must have sucked in quality.

[3467]

Read the report. The data was impressive. One of the most depression reforms was better economic data though we have Economic Censuses that were pretty good back to 1700.
What the could not have predicted is how much the system would determine where future development would and did occur.
What I find interesting is how we never reached a post interstate consensus to fill those gaps.
We discussed how ISTEA tried . The plans we state by state and varied from grandiose like the Illinois supplemental freeway system  Iowa Missouri Pennsylvania and New York had similar plans but most states didnt.

[Rothman]

Economic data is one thing, but I doubt their traffic data was that comprehensive.

[thspfc]

Terrain and wealth were probably major factors alongside population.

[Hot Rod Hootenanny]

Economic data is one thing, but I doubt their traffic data was that comprehensive.

My guess is that if Ohio State University has a copy of a 1925 traffic study for New York, that an employee of NYSDOT could find the same study as well.

Link to OSU's copy: https://osu.on.worldcat.org/search?databaseList=1723%2C3909%2C239%2C1461%2C1708%2C638%2C1725&queryString=ti%3AHighway+traffic+study#/oclc/17335379

[Rothman]

Reminds me of when they tried predicting the weather by radioing the next station up the line.

[3467]

The study I saw was post interstate and showed freeways and expressways for I recall US 11 and 219 and something through the finger lakes....

[The High Plains Traveler]

Kind of next-step to the original process, but I find myself wondering when the general alignments for the Interstate routes were decided. I look at I-70 crossing Kansas, and from east to west it generally follows U.S. 40, hitting Topeka, Salina, Russell, Hays, and WaKeeney. But then, it swings northward to U.S. 24 and catches Colby and Goodland, and then Burlington CO. I wonder whether this routing, and the point at which it swings north, was a source of controversy. Then, there's I-35 from Iowa to Minnesota. As late as 1964, there is a map showing its proposed route following U.S. 69 all the way from Des Moines to Albert Lea, MN. The final route swings east closer to U.S. 65 so it can serve Mason City, IA.

[sprjus4]

Kind of next-step to the original process, but I find myself wondering when the general alignments for the Interstate routes were decided. I look at I-70 crossing Kansas, and from east to west it generally follows U.S. 40, hitting Topeka, Salina, Russell, Hays, and WaKeeney. But then, it swings northward to U.S. 24 and catches Colby and Goodland, and then Burlington CO. I wonder whether this routing, and the point at which it swings north, was a source of controversy. Then, there's I-35 from Iowa to Minnesota. As late as 1964, there is a map showing its proposed route following U.S. 69 all the way from Des Moines to Albert Lea, MN. The final route swings east closer to U.S. 65 so it can serve Mason City, IA.

In Virginia, there was a debate on where to build I-64 between Richmond and Lexington in the 1950s - 1960s. One routing followed US-460 between Richmond and Roanoke serving Lynchburg, and another routing followed US-250 between Richmond and Staunton serving Charlottesville. Ultimately, the latter was constructed, and the former, US-460, was mostly widened to a 4-lane divided highway. As a result, Lynchburg remains the largest city in the state not served by an interstate highway. Has the southern routing serving Lynchburg been constructed for I-64, Charlottesville would likely be the largest city in the state not served by an interstate highway.

[ilpt4u]

Kind of next-step to the original process, but I find myself wondering when the general alignments for the Interstate routes were decided...

Maps were printed and construction had already began on the IL US 50/IN US 150 routing for I-64 between St Louis and Louisville (Freeway section that begins in Lawrenceville, IL and continues around Vincennes, IN, for example)...but a Mount Vernon, IL and Evansville, IN lobbying group successfully lobbied to move the ultimate routing to the route that it is today, roughly following old US 460

[Beltway]

Kind of next-step to the original process, but I find myself wondering when the general alignments for the Interstate routes were decided. 

In Virginia, there was a debate on where to build I-64 between Richmond and Lexington in the 1950s - 1960s. One routing followed US-460 between Richmond and Roanoke serving Lynchburg, and another routing followed US-250 between Richmond and Staunton serving Charlottesville. Ultimately, the latter was constructed, and the former, US-460, was mostly widened to a 4-lane divided highway. As a result, Lynchburg remains the largest city in the state not served by an interstate highway. Has the southern routing serving Lynchburg been constructed for I-64, Charlottesville would likely be the largest city in the state not served by an interstate highway.

The Virginia Interstate system was mapped out in 1956 and 1957, preliminary engineering determined the proposed routes then.  I have seen the document at the Virginia Transportation Research Council (VTRC) library.

Very little changed other than that part of I-64.  BTW, the route between Richmond and Roanoke is US-360 and US-460, and all of that is 4 lanes.  The northern route also was originally planned to pass just north of Charlottesville, and the section between Lexington and Clifton Forge was a direct route with a mountain tunnel.

[J N Winkler]

Kind of next-step to the original process, but I find myself wondering when the general alignments for the Interstate routes were decided. I look at I-70 crossing Kansas, and from east to west it generally follows U.S. 40, hitting Topeka, Salina, Russell, Hays, and WaKeeney. But then, it swings northward to U.S. 24 and catches Colby and Goodland, and then Burlington CO. I wonder whether this routing, and the point at which it swings north, was a source of controversy.

I haven't looked in primary or secondary sources, but I suspect it was uncontroversial because the adopted routing is a no-brainer in terms of being more direct (US 40 swings a bit south before it swings back up north) and serving cities of larger population in 1940.  Hugo, Kit Carson, Cheyenne Wells, and Sharon Springs all had sub-1000 populations each, while Colby at the time already had 2,500 people and Burlington 1,300.  (The aggregate populations of Colby and Burlington alone are now about 10,000, while the US 40 cities have experienced very little population growth in the last 70 years.)

In Kansas, however, there were definitely Interstate routing controversies.  In another thread we have discussed BPR's desire to incorporate the entirety of the Kansas Turnpike as the I-35 routing:  I-135 and free I-35 between Emporia and KC, along with deletion of planned free I-35 touching Wichita, Newton, and Emporia, were part of the resulting compromise.

There was also an unsuccessful attempt to locate I-70 in the more heavily populated Kaw River valley (mostly US 24) between Topeka and Junction City.  IIRC, Army influence (Fort Riley is near the western end) was decisive in securing the present routing along what was then the US 40 "beeline."  In the 1940's Manhattan, the largest of the cities bypassed at distance, had a population of about 12,000.  Now it has 54,000 people and two high-capacity connections to I-70:  the K-18 freeway (built late 2000's/early 2010's) in the west and the K-177 expressway (built mid-1990's) in the east.

Then, there's I-35 from Iowa to Minnesota. As late as 1964, there is a map showing its proposed route following U.S. 69 all the way from Des Moines to Albert Lea, MN. The final route swings east closer to U.S. 65 so it can serve Mason City, IA.

That is closer service to a town with 1960 population of 30,000 (it has since declined slightly), the largest across a fairly wide region in northern Iowa, at no length disadvantage compared to a routing sticking more closely to the US 69 corridor, which itself bends to the east closer to Albert Lea.

[wxfree]

They apparently asked states for their ideas early on.  This Texas Highway Commission minute order from February 26, 1945 directs the state highway engineer to submit the following roads to the Public Roads Administration for inclusion in the Interstate highway system.

1. U. S. Highway 66 from the New Mexico State Line to the Oklahoma State Line.

2. U. S. Highway 80 from the New Mexico State Line north of El Paso to the Louisiana State line east of Marshall.

3. U. S. Highway 67 from Dallas to the Arkansas State Line at Texarkana.

4. U. S. Highway 290 from its junction with U. S. Highway 80 to its junction with State Highway 27 northwest of Mountain Home; State Highway 27 from its junction with U. S. Highway 290 northwest of Mountain Home to its junction with U. S. Highway 87 at Comfort; U. S. Highway 87 from its junction with State Highway 27 at Comfort to San Antonio; and U. S. Highway 90 from San Antonio to the Louisiana State line east of Beaumont.

5. U. S. Highway 81 from its southern terminus at Laredo to its junction with U. S. Highway 77 at Hillsboro; U. S. 77 from its junction with U.S. 81 at Hillsboro to the Oklahoma State line north of Gainesville.

6. U. S. Highway 81 from Hillsboro to Fort Worth and U. S. Highway 377 from Fort Worth to its junction with U. S. 77 at Denton.

7. U. S. Highway 75 from its southern terminus at Galveston to Dallas.

This is Minute Order 21014, starting on page 9.

https://publicdocs.txdot.gov/minord/MinuteOrderDocLib/003676343.pdf

[sprjus4]

BTW, the route between Richmond and Roanoke is US-360 and US-460, and all of that is 4 lanes.

The shortest route, the arterial routing between Richmond and Roanoke still has a 10 mile stretch still at two lanes, the stretch of VA-307 linking US-360 and US-460.

The quickest route between Richmond and Roanoke however would be I-64 and I-81, despite the additional mileage. Google estimates its approximately 16 minutes quicker, 25 additional miles. I've traveled that route quite a few times in the past, and I've always utilized the interstate. I've only ever used US-460 between Lynchburg and Roanoke, and that was because I was originating in Lynchburg.

[jeffandnicole]

Data in 1944 must have sucked in quality.

Economic data is one thing, but I doubt their traffic data was that comprehensive.

Any data to prove your conclusion, or are you relying on "it was 1944 so they didn't have computers and didn't know anything".

[Rothman]

Data in 1944 must have sucked in quality.

Economic data is one thing, but I doubt their traffic data was that comprehensive.

Any data to prove your conclusion, or are you relying on "it was 1944 so they didn't have computers and didn't know anything".

More along the lines that most DOTs didn't have a comprehensive traffic count program up and running in 1944, even with the available technology at the time.

Heck, some DOTs didn't even exist in 1944 and states were more focused on capital projects and maintenance rather than traffic operations and management.

Even then, I do doubt that if I threw the level of traffic data that we have available now at a 1944 DOT that their room full of clerks with slide rules would have the same processing power we do today.

[J N Winkler]

More along the lines that most DOTs didn't have a comprehensive traffic count program up and running in 1944, even with the available technology at the time.

Heck, some DOTs didn't even exist in 1944 and states were more focused on capital projects and maintenance rather than traffic operations and management.

I don't think that is completely true.  Yes, the 1940's are early for automatic traffic count stations, but the statewide highway planning surveys were a big thing in the 1930's (establishing the data basis for the toll roads/free roads survey and the 1944 interregional highways report), and taking spot counts of traffic was an established practice by the 1920's.  The 1930's were also when state DOTs (or, to be precise, their predecessor agencies) got heavily into traffic operations, with the personnel responsible initially operating out of their maintenance units.

[Beltway]

BTW, the route between Richmond and Roanoke is US-360 and US-460, and all of that is 4 lanes.

The shortest route, the arterial routing between Richmond and Roanoke still has a 10 mile stretch still at two lanes, the stretch of VA-307 linking US-360 and US-460.

That saves 5 miles, and there is no need to use that, particularly a large truck would be more suited to use US-360 and US-460.

VA-307 is an ok 2-lane highway, but not a "high type" 2-lane highway.

The quickest route between Richmond and Roanoke however would be I-64 and I-81, despite the additional mileage. Google estimates its approximately 16 minutes quicker, 25 additional miles. I've traveled that route quite a few times in the past, and I've always utilized the interstate. I've only ever used US-460 between Lynchburg and Roanoke, and that was because I was originating in Lynchburg.

If one is already on I-64 to the east of Richmond or east of Williamsburg, it might seem more natural to stay on I-64 and maybe use I-295 as well to bypass to the north.  Smooth routing compared to having to use segments of I-64, I-95, VA-195, VA-76 and VA-288, to get to US-360 west of Richmond; it's all freeway but you need to make all those connections.

If you live in or near Richmond, then US-360 and US-460 tends to be a visible and attractive route to Roanoke.  Google just estimated I-64/I-81 is approximately 14 minutes less, 25 additional miles.  That assumes no delays on I-81.  US-360 and US-460 is a reliable and more peaceful drive and it is 25 miles shorter.  I would head out VA-76 and VA-288 to reach US-360.

[Beltway]

Even then, I do doubt that if I threw the level of traffic data that we have available now at a 1944 DOT that their room full of clerks with slide rules would have the same processing power we do today.

You always hear about these slide rules.

Mechanical calculators existed long before the 1940s.

The B-29 Superfortress had analog computers for the remote control of the defensive gun turrets in 1944, and they were far more effective than the older style manual turrets.

[3467]

Illinois was doing traffic counts that far back. The first Interstate map appeared in the 1947 annual report and pretty much followed the original map except for the previously mention move of 64.
I have heard stories about other routings  but they are just that stories.

[Rothman]

More along the lines that most DOTs didn't have a comprehensive traffic count program up and running in 1944, even with the available technology at the time.

Heck, some DOTs didn't even exist in 1944 and states were more focused on capital projects and maintenance rather than traffic operations and management.

I don't think that is completely true.  Yes, the 1940's are early for automatic traffic count stations, but the statewide highway planning surveys were a big thing in the 1930's (establishing the data basis for the toll roads/free roads survey and the 1944 interregional highways report), and taking spot counts of traffic was an established practice by the 1920's.  The 1930's were also when state DOTs (or, to be precise, their predecessor agencies) got heavily into traffic operations, with the personnel responsible initially operating out of their maintenance units.

Sure, but the datasets available now dwarf anything they could collect back then and, even if they could, they wouldn't be able to analyze it in a very timely manner.

People have been collecting weather data for millennia and yet our forecasts didn't approach anything meaningful until the 20th Century.  Intention does not equate capability or quality.

[sprjus4]

That saves 5 miles, and there is no need to use that, particularly a large truck would be more suited to use US-360 and US-460.

VA-307 is an ok 2-lane highway, but not a "high type" 2-lane highway.

VDOT signs VA-307 as "To US-460 West; Farmville / Lynchburg" on US-360 heading westbound, and simply "VA-307 East; Richmond" on US-360 heading eastbound.

For a trucker, taking US-360 to US-460 would add 7.5 miles, which you've pointed out in the past mileage for truckers is strict.

VA-307 is a suitable road for trucks, and is mostly straight and has rolling terrain. VDOT wouldn't sign it as the main route between the two points if it couldn't handle truck traffic.