Amtrak Joe, Hometowns, and what Amtrak’s Past Can Teach Us About its Future
I read a lot about how and why passenger rail or high-speed rail can’t or won’t work in the United States. There is a lot of useful history embedded in those arguments — certainly kernels of truth. But I think there are some serious misconceptions out there about what the railroad history in the United States has been — and what it can be in the future.
This is an essay about the president’s hometown, my hometown, our adopted hometowns, and how our unique American history has resulted in some highly divergent transportation outcomes in different cities as well as what I think our nation is capable of in the future.
I’m from Kalamazoo, Michigan, a smallish city with a cool name halfway between Detroit and Chicago. Hopefully everybody knows that Kalamazoo is a real place. I’m sure “everybody” knows that President “Amtrak” Joe Biden is from Scranton, Pennsylvania.
Fewer know that Kalamazoo is served by Amtrak-owned tracks that have hosted 110 mph trains for more than nine years. Just as few know that Scranton was once served by one of the earliest high-speed rail lines in world history…but that it has been derelict for more than 40 years.
High Speed Rail History
It seems to me that the popular understanding of the history of high speed rail starts with Shinkansen (Japan) in the mid-60s, goes to France for the TGV in late 70s, spreads slowly around the world for twenty years (including a “failed” attempt in the US with the Acela) and then accelerated globally over the last ten years.
I’ll make the case that is wrong. The history started in the United States in the early 20th century. Major railroads such as the New York Central and Pennsylvania Railroad improved their tracks and their rights-of-way. One, the Lackawanna (officially the Delaware, Lackawanna and Western Railroad), opened an engineering marvel in 1911…But more on that later.
Train technology followed suite. Special-built trainsets such as the M-10000 and the Burlington Zephyr were developed that easily sped past 100 mph and averaged end to end trip speeds over 75 mph for a thousand miles (including stops). This was at a time when a late-model car such as a Ford Model A couldn’t reach over 70 mph. Nationwide, railroad “streamlined” trains. Much of this was marketing hype (art deco industrial design and fantastical names), but it was built on technological and operating improvements. Technology included advanced internal-combustion engines and light-weight materials. Operating practices included precision scheduling and services optimized to serve the biggest travel markets at the most profitable times.
By the late 30s, most Americans were served by some of the most technologically advanced railroads in the world. Most of these services improved until average trip speeds were around 60 mph over a distance of 300 or more miles in the 40s, when progress stalled during the war effort.
The United States Takes a Different Path
But travel wasn’t safe at that time. Automobile drivers frequently killed pedestrians as well as themselves and their passengers, and trains crashed all the time. My late grandfather (1913–2007) used to tell me stories about interurbans scaring horses and killing people. In 1947, after a disastrous train crash killed 45 people, a well-meaning rule requiring more advanced signaling technology in order for trains to operate over 80 mph was passed at basically the exact same time that:
- the Federal Government was ramping up subsidized construction of the interstate highway system with the Federal-Aid Highway Acts of 1944 and 1952, and
- technology and know-how spin-off from the war effort accelerated civil aviation.
We could wonder what American transportation history would have been if this rule were in place at the beginning of the streamliner revolution in the early 30s or if it had been put into place along with post-war economic stimulus to install the technology on the nation’s railroad. Unfortunately, we know the result was that most railroads slowed their trains down and began a long decline and contraction of their services in the face of subsidized competition.
Skip ahead to 1970. Railroads are consolidating and going bankrupt left and right. Passenger trains are being cut. In most cases, railroads were forced to operate passenger trains as a part of a regulation regime due to their apparent status as a monopoly (and also perhaps, because the public often subsidized the original creation of the lines). Especially in the industrial heart of the country in the Northeast and the Midwest, there were so many competing parallel rail lines, that the only way railroads could continue operating was to merge with a competitor, abandon excess infrastructure, salvage parts, and run what seemed like the best pieces of the resulting enterprise.
The New York Central (NYC) and the Pennsylvania Railroad (PRR), primary competitors between New York and Chicago, as well as most every city in between, merged in 1968 and went bankrupt in 1970. (Imagine if Google and Facebook merge in 2027 but then go bankrupt in 2029!) Their other competitors were largely doing the same.
Amtrak was created in 1970 to take over passenger train obligations in 1971. This article is a great history on the legislation, and this article is a great history on its implementation. A few years later, the federal government created Conrail to allow for a rationalized network of rail lines in the Northeast and Great Lakes. Elsewhere in the Midwest other railroads were similarly merging into their competitors and fading away toward the same end.
The Path Not Taken
In my opinion, the biggest failing of the Rail Passenger Service Act of 1970 is that it didn’t preserve tracks for passenger trains, even at a time when the freight rail system needed to cut excess lines from its inventory.
Railroads are their own animals, but they are can be analogous to streets and highways. Freight trains are long, heavy, and (relatively) slow. Passenger trains are short, light, and (relatively) fast. Mixing them reduces capacity. Imagine a highway filled with cars where everybody is in a car (or pickup, this is ‘Merica) going 5 mph over the speed limit (again: ‘Merica). A lot of cars can fit on that highway. It might take more imagination unless you’re a trucker, but that same highway can hold a lot of semi-trucks going 5 mph slower. But if you try to put a couple of fast cars onto that highway, they aren’t going 5 mph over anymore, and you’re not fitting as many cars or semis. (The same analogy works with a bike lane and a car lane side-by-side compared to a bunch of bikes mixed into two general lanes.)
The biggest railroads in the US figured that fact out way before 1970s. Both PRR and NYC had four-track rights-of-way for the busiest parts of their lines that also supported passenger trains. The Japanese and French had figured the same thing out as well when they maxed out what they could do on their old lines and realized they could move a lot more passengers more quickly on a section of separate bypass track.
It seems to me that we collectively forgot this when Amtrak was in its infancy. I believe that had the Amtrak/Conrail route rationalization process resulted in a network of tracks for passenger trains east of the Rocky Mountains, we’d be in a whole different place with respect to passenger rail at this time. That system may have also resulted in an open-access network for lightweight and high-value parcel freight — but that’s another idea.
Amtrak came to own most of the Northeast Corridor in the mid-70s, but many of the lines that could have been separate routes for passenger trains from the lines that got merged into CSX or Norfolk Southern (in the east) and BNSF or Union Pacific (in the west) were either abandoned or were used for short line connecting railroads that have little use for high-quality infrastructure. At the same time, four track railroads had the “excess” (freight) capacity removed.
Just a few examples:
- Wabash line from Toledo to Ft Wayne: Part abandoned and part shortline.
- Erie line into Chicago: Abandoned.
- Baltimore and Ohio mainline through WV: Abandoned.
- NYC four-track line through Upstate NY: Mostly two tracks.
- PRR four-track line through Central PA: Mostly two-three Tracks.
Which brings me back to Scranton (Biden) and Kalamazoo (Myers) as well as Wilmington, DE (Biden) and Raleigh (Myers)…
Amtrak came to own the line between Kalamazoo and Porter, Indiana (98 miles) at about the same time that they got most of the Northeast Corridor (NEC) in the mid-70s. At that time, Amtrak was the only user of the line; Conrail was routing freight trains different ways. Similarly, commuter trains and Amtrak were the primary users of the NEC. Most freight was either perpendicular to it, or along parallel routes (now controlled by CSX).
Amtrak and the state of Michigan have slowly and steadily improved the Michigan service. In 2005, that technology was implemented to increased speeds on this section to 95 mph. In 2012, it was increased to 110 mph. In 2012, the state of Michigan bought the rest of the line from Kalamazoo to Metro Detroit. Since then, with limited funding, they’ve repaired the track and cut 30 minutes out of the schedule. Reports are that this spring an additional 22 miles will go from 79 to 110 mph. Later in 2021, another 48 miles will be up to 110 mph and then in 2022 the last 63 miles will be running with a 110 mph top speed. That’s a total of 231 miles.
Not that every curve is so fast, but the average speed between Michigan City and Kalamazoo is currently 79 mph. The current Acela covers Washington to New York at an average speed of 82 mph. Even one of the famed TGV’s newest lines, with top speeds of 220 mph, averages much less (about 150 mph) on a trip from Paris to Strasbourg. Many TGV services are slower when reaching destinations not on the new high-speed lines, reducing average trip speed. A second point of global comparison: A German InterCity Express trip between Koln and Hannover averages around 70 mph. I’m pretty tired of reading comparisons between Acela’s average speed and some other country’s train’s top speed. It’s a meaningless comparison.
If Kalamazoo to Dearborn becomes almost as fast as Michigan City to Kalamazoo (it’s curvier), they can cut an extra 30 minutes out of the schedule next year and Detroit to Chicago will take about 4.5 hours next year. This time is the same as driving aggressively and without stopping — but a whole lot better for the environment (not to mention with an option for beer). My point is not that Amtrak and Michigan are operating world-class high-speed rail, but rather that they are running good transportation and that it can and should be expanded.
The challenge Detroit-Chicago trains have is congestion and slow speeds for the last 40 miles between Porter, Indiana and Chicago Union Station. This fact limits reliability, caps the frequency of trains, and slows down the total trip time (Even a modestly faster 60 mph average through this section would save 15–20 minutes).
Of course, Wilmington, DE is on the northeast corridor. We’ve all heard about how Senator Biden used it to commute home after conducting Senate business. Stopping at Wilmington, between Washington and New York, there are about 25 round trips of regional and long-distance trains (average speed about 65 mph) and a ten additional Acela round trips (average speed about 80 mph).
The NEC wasn’t particularly unique before the Northeast Corridor Improvement Project (NECIP), which started in 1976, and the electrification project in New England during the 1990s. Most any other non-mountain railroad can be upgraded to provide average trip speeds around 70–80 mph. The NEC can be challenging because there are so many trains. In addition to the ~33 Amtrak trains at Wilmington, there are hundreds of commuter trains at various points along the corridor.
They key to the NEC and the Michigan line is passenger-first track. In the case of Michigan, the current lack of passenger-first track for 40 miles caps the service to four round trips per day. And remember, frequency is freedom!
But what about Scranton and Raleigh?
The Lackawanna Cutoff is a rail line from Northern New Jersey to the Delaware Water Gap. Along with the Nicholson Cutoff between Scranton and Binghamton, NY it created the early 20th Century version of a TGV high-speed line from the Hudson River to Scranton and Binghamton. There are huge viaducts as well as cuts, fills, and tunnels to make a straight path for trains through hilly and terrain. It’s not unlike some of the French TGV infrastructure.
In 1979, right after Amtrak got 98 miles of the Michigan Line and the NEC, the Lackawanna Cutoff was abandoned. A train hasn’t run on it my entire life. Scranton has no Amtrak service, and certainly not an upgraded line with ~70 mph average service, taking travelers to Manhattan in about 2 hours. This line is now owned by the state of New Jersey, which has plans to return it to service.
Ten years later, the direct rail route between Raleigh and Richmond Virginia lost its rails. Amtrak got bumped to a route 40 miles longer and busy with freight trains when two competing railroads merged and then pruned their route structure. Rather than consolidating freight trains onto the better line for freight and creating a dedicated passenger line, we consolidated freight and then crammed passenger trains onto the same line!
The states of Virginia and North Carolina have been working on that problem for years. Four years ago, they completed the environmental study to return the line to service and to upgrade it to faster speeds. This week, Virginia finalized its deal with CSX to purchase the line, along with half of the right-of-way between Richmond and Washington. North Carolina has announced a similar deal.
The future of passenger rail in the United States is best built on dedicated passenger track. This is how the NEC came to be, how Michigan laid the foundation of a good service to my hometown of Kalamazoo in the 70s and has built on it since 2012. It is how NC and VA are working to bring direct fast trains to my adopted hometown of Raleigh. It should have been how the Lackawanna cutoff was converted to a passenger line in the 70s so that Amtrak Joe could have taken the train home for a visit. (It may be what happens in the future.) There is easily enough underutilized railroad ROW to connect the NEC to Chicago through cities like Pittsburgh, Youngstown, Buffalo, Cleveland, Toledo, and Ft. Wayne. There are opportunities to extend this network into the Great Plains.
Some people will choose to take a train that takes 16 hours New York to Chicago (the historical streamliner schedule) just as some choose to take that ~20 hour trip today on the congested freight-first network. Even more would take it if the average speed is brought up to the standards of Amtrak Michigan or Acela (12–13 hours). Mostly, this network will serve trips between the innumerable intermediate cities and towns, where the only current alternative is driving.
New high-speed lines can be built around the slowest or most congested sections of this network. In places further west where railroads didn’t quite develop as many parallel and competing lines before they met subsidized competition, there are fewer opportunities to repurpose one line for fast passenger trains while one line remains for heavy freight. But this backbone is the foundation that meme-map fantasies for high speed rail need to become goals or plans.
