Passengers Are Asking Freights the Wrong Questions

Is it more cost-effective in the long run to go a different direction?

I won’t get into why specifically, but this week I’ve been wrestling with a kind of fundamental question when we start to examine the delicate dance we do with Class I railroads to expand passenger service. Nearly since Amtrak was created in 1971, we’ve approached it the same way: “What do we have to build for the freight railroad before it will let us run another passenger train?” But maybe…we’re asking the wrong question?

I’ve heard versions of the same explanation from freight railroads for years; in fact, I just heard it again yesterday afternoon in a meeting I was in. A host railroad facing a request for new or additional passenger service goes through some version of the following process – start with the freight network as it exists. Assume it is operating efficiently (a truly heavy assumption given the empirical data, but let’s just assume it here for now). Add the proposed passenger train to a computer model, an industry standard called Rail Traffic Controller. Measure the disruption. Find the chokepoints. Then calculate the sidings, crossovers, additional main tracks, yard improvements, and other infrastructure needed to restore the freight railroad to its starting condition.

I don’t really dispute this on its face; there is a certain logic to that rubric. But there’s also a gigantic public-policy assumption buried inside it. The assumption? Freight’s existing operating environment is fixed, while the passenger train is the intruder. Therefore, the public’s job is to spend whatever taxpayer dollars are necessary to insert the passenger train without disturbing the freight railroad…despite the fact that Amtrak’s very creation represented a taxpayer-funded bailout of the Class Is losing money operating passenger trains.

The Federal Railroad Administration has actually identified this issue before. In a report to Congress on shared railroad rights-of-way, the agency noted Amtrak’s concern about modeling based on “zero impact” rather than “unreasonable impairment” of freight operations, which is the regulatory standard, and that makes a truly important difference. (Also, full disclosure, we at Rail Passengers contributed to that FRA report.)

Photo by Jared Patrick on Unsplash

Just consider the two decades of fist-fighting and litigation it took to restore passenger trains between New Orleans and Mobile.

In 2017, CSX estimated that infrastructure needed for Gulf Coast passenger service would cost $2.3 billion. A congressionally directed federal study produced a figure around $120 million. By late 2021, CSX and Norfolk Southern were seeking a little over $400 million. Then-House Transportation & Infrastructure Chairman Peter DeFazio put all three figures on the record before the Surface Transportation Board in 2022.

Now, just for the moment leave aside which number was correct, and instead ask another question: “What else could we have bought for $2 billion?” (And yes, I know we didn’t wind up spending $2 billion to coax the Amtrak Mardi Gras into life but follow me through on this and keep in mind the opportunity costs, too.)

Our rough look at Federal transit construction data suggests that ordinary new at-grade guideway and track can be built for something on the order of $10 million per mile in favorable circumstances, before major stations, extraordinary structures, and difficult land acquisition. At that scale, $2 billion begins to resemble the cost of 200 miles of new railroad. Not another siding owned by CSX, or another crossover designed principally to keep Norfolk Southern whole. A railroad, our railroad, with track designed around passenger trains and passenger signaling and passenger dispatching. Infrastructure designed for 110 mph operation where geography permits it. And, to my mind one of the most important benefits, we get an asset the public still owns after the check clears.

That suggests a simple change in federal policy.

Whenever a freight railroad says substantial capital investment is needed to accommodate new or expanded passenger service, FRA should be required to perform what I’ll call a “Passenger Rail Infrastructure Alternatives Test.” If you want, call it the “Build-or-Buy Test.”

Alternative One asks what we ask today: what must we invest in the host railroad to operate the service? But Alternative Two asks something we rarely ask: what would it cost to acquire an adjacent corridor and build infrastructure intended for passenger trains?

You could compare them across 30 years. You would include maintenance, any host payments in the mix, future capacity demands, reliability, the ability to add more trains later, and then who owns the asset when we’re finished.

Then decide.

Now, I’ll be the first to say that sometimes, maybe even often, the freight railroad will win that comparison easily. A $30 million siding is a much better deal than building 70 miles of new railroad through difficult terrain. Fine. Build the siding.

But suppose the answer is $800 million of freight-railroad improvements to launch six passenger roundtrips – or $1 billion for a dedicated passenger track that could eventually handle twelve or twenty? That ought to be a very different conversation.

We already have examples of states beginning to think this way, and I was reminded about them during my meeting yesterday afternoon.

The separation itself has real economic value, including value to the aggrieved Class Is who would really prefer to see passenger rail just evaporate off their territories. Virginia paid CSX $525 million for a property transaction, acquiring 384 miles of right-of-way and 223 miles of track, including half of the CSX-owned right-of-way between Washington and Petersburg. The Virginia Passenger Rail Authority explicitly says the program is intended ultimately to achieve separation of passenger and freight operations, which is honestly what both freight and passenger really want, right?

And Utah is preserving a corridor alongside Union Pacific so that future passenger trains don’t have to depend indefinitely on shared freight tracks, so a much smaller scale but the same principle at work.

Nobody calls either of those ideas crazy, they just look like infrastructure planning.

So maybe we should apply the same logic nationally?

Most of my conversations with Class I executives, even those sympathetic to passenger rail, offer some variation of the (very real and legitimate) pressure they face to produce shareholder value and operating efficiency. They have a business to run. I get it. (They also have a statutory obligation they took on in 1971 as a quid pro quo for their taxpayer-funded rescue, but I’ll set that aside right now.) They want to have a business conversation. So, let’s have it.

We should deliberately make speed and consensual separation part of the thing we’re purchasing from the Class I, if we do something like the Virginia CSX deal. We should attach value to what the Class I gives up besides land, along with putting a value on what we cease demanding from the railroad.

Imagine gliding into a glassy high-rise boardroom and making this offer to NS:

“We will pay you generously for a permanent 75-foot passenger transportation corridor alongside your main line…say, $83,000 per acre. You retain your railroad. We build ours. We will assume maintenance, inspection, signaling and dispatching of our infrastructure. We will design our own capacity for future passenger growth. We control our own liability and indemnification. And except where physical interfaces are truly unavoidable, future increases in passenger service will no longer require capacity negotiations with you.”

No more haggling, no more expensive white-shoe litigation over what “preference” means or “impairment,” and no dark comments about confiscation or “takings” or the like. Just a business proposition.

Don’t jump on me in your comments below, I’m not suggesting that Congress write a check tomorrow for 22,000 miles of parallel Amtrak railroad all across the country (at least not yet!).

But what I am suggesting is that before taxpayers put another billion dollars into somebody else’s railroad merely to preserve a pretty unimpressive operating status quo, we should be required to ask what that billion dollars could build for passengers instead. Do that honestly, corridor after corridor, and let’s see where the answer leads us, taking into account what it costs to do it right plus the additional costs of doing nothing. We might discover that sharing is still the cheapest solution in most places. Or we might discover that after decades of paying to fit passenger trains into somebody else’s railroad making nobody happy, the most-effective, most-efficient, highest-and-best-use long-term answer is astonishingly simple: Build our own.

What do you think?