Want to Cut Passenger Train Costs? Run More Trains!

By Fritz Plous

When passenger-train critics go after their favorite target they use their favorite argument—so-called “high costs per passenger.”

“The federal government is paying $300 in subsidies for every passenger on the Sunset Limited!” — the anti-train crowd wails. “Why, for that kind of money we could buy every passenger on that train a plane ticket.”

Forget for a moment that most city pairs served by the Sunset don’t have air service. The real question: is loss-per-passenger a meaningful metric?

The answer is no — and for several reasons.

Transportation “losses” ignore value

First, loss-per-passenger is meaningless when taken out of context because it does not identify value — especially value to the economy, which benefits whenever an individual travels.

Look at all that real-estate development around O’Hare: The airport itself loses money handling airplanes, and the Federal Aviation Administration requires huge subsidies to run the Air Traffic Control system and enforce safety regulations.

But government recovers all of those losses — and earns billions more — in the taxes paid by individuals and businesses that use air transportation to create new wealth. A “money-losing” government transportation activity — an airport or a highway — can be a powerful driver of business growth. In fact, that’s why the government subsidizes its transportation infrastructure. A “money-losing” passenger-train network can perform the same economic magic if given the same resources and rules.

Charging losses to each train masks the real problem — and the real opportunities.

But failure to calculate value is only one reason why loss-per-passenger-pertrain is a silly number. The real absurdity is assigning costs to individual trains rather than to the overall system — the business that runs all the trains.

Analyze this: Imagine that Amtrak’s Texas Eagle is losing $100 per passenger on its 1,306-mile itinerary between Chicago and San Antonio.

Now imagine the Texas Eagle has been broken up into three trains — an Illinois Eagle operating the 284 miles between Chicago and St. Louis, an overnight Ozark Eagle covering the 707 miles between St. Louis and Dallas via Little Rock, and a Texas Eagle day train serving the last 315 miles between Dallas and San Antonio. The equipment would continue to operate from Chicago through to San Antonio in order to minimize operating costs.

Bingo! — the Eagle just got cheaper. Instead of one “unacceptable” train costing the government $100 per passenger, we now have three “acceptable” trains each costing the government $33 per passenger. Problem solved.

If it was ever a problem to begin with.

Losses can be cut if investment is increased

Assigning costs to each passenger on each individual train leads to another absurdity. It forces certain kinds of trains, particularly long distance trains that operate only once a day or less, to absorb an unfairly high share of the carrier’s overhead costs.

As a rule, overhead costs, such as administration, station expenses, building rental, and utilities, are relatively fixed. As business grows, overhead expenses stay at pretty much the same level, so that each new customer removes more and more of the overhead burden from all the others.

For example, if it costs $500 per day to keep a station open and the station handles 500 passengers per day, a dollar of each passenger’s fare must be allocated to the costs of that station.

But if the railroad can build traffic to 1,000 passengers a day, the same station can process each passenger for 50 cents. The cost per passenger drops as the number of passengers rises. Voila! The railroad station just doubled its productivity.

But it’s very hard to raise the productivity of stations serving long-distance trains, since doubling the number of passengers at most stations would require doubling the number of seats. Amtrak has no budget to acquire the additional cars and locomotives needed to run additional trains or to make the existing trains longer.

That means Amtrak is barred from increasing its productivity. Amtrak can’t do what the California Department of Transportation did when it used funds from a 1990 bond issue to buy new cars and locomotives for Amtrak’s California intrastate routes. Amtrak is stuck.

Go West and cut your cost per passenger!

What California learned is that when it ran more trains it reduced its cost per passenger. When the Capitol Corridor trains were inaugurated between Oakland and Sacramento in 1991, costs were high because the service was in startup mode. Only three round trips per day were operated, and in the first full year of operation the subsidy per passenger came to $22.45.

But by the 1994-95 fiscal year, annual ridership on the three trains had nearly tripled, and the loss-per-passenger dropped to $16.96. In April of 1996 a fourth round trip was added. That generated additional expense, but it also made the Corridor more attractive. Ridership jumped more than 13 per cent, and the loss-per-passenger sank to $15.56.

The economic principle for trains is no different than for a restaurant, a manufacturer or a retailer: The more hamburgers, widgets or shoe sales you can generate under the same roof, the less it costs to produce each one. In MBA language, unit costs drop as volume increases.

So what’s the right number? Farebox recovery.

If loss-per-passenger-per-train misrepresents passenger-train performance, what’s the right number?

It’s something called “farebox recovery,” the portion of the business’s total operating costs that is borne by all its customers. Farebox recovery is the standard measurement of productivity in the transit industry. Chicago’s Metra commuter rail system, for example, recovers about 55 per cent of its operating costs from passenger fares. That’s considered strong performance for a mass-transit operation. Amtrak recovers about 65 per cent of its operating costs from passenger fares.

And what about Amtrak’s two competitors, the government-funded civil-aviation infrastructure and its manager, the Federal Aviation Administration; and the government funded highway network and the local streets that feed it? How much of their costs do they recover from their users?

A 1989 study by the Congressional Budget Office showed that the 15 per-cent federal airline ticket tax covered only about two-thirds of the government’s civil-aviation costs. That’s about a 65% recovery — the same as Amtrak. The balance was covered by subsidies from the General Fund (i.e., income taxes).

A 1992 study by Chicago’s Metropolitan Planning Commission found that highway costs in the six-county Northeastern Illinois region had about the same recovery rate: motor-fuel taxes, license fees and taxes on the sale of motor vehicles covered about two-thirds of direct highway costs. The balance had to be subsidized from non-user sources such as sales taxes, real-estate taxes and income taxes. Similar studies have shown about the same 65 percent highway “farebox ratio” nationwide.

However, most studies of civil aviation and highway financing contain a caveat: the two-thirds of costs ostensibly covered by user fees represent only the direct costs of building and maintaining the infrastructure. A major share of highways costs, including law enforcement and snow plowing, are paid out of police, fire and public-safety budgets funded by such non-user sources as property, income and sales taxes.

Nor do studies of aviation and motoring calculate the costs of the pollution both modes generate, or the huge military budget needed to supply the massive volumes of “cheap” Middle Eastern oil on which both forms of travel critically depend. Those costs are shed to other areas of government or simply to the economy as a whole. When these hidden costs are included as part of the real price of travel, driving and flying turn out to have no better farebox-recovery ratios than passenger trains or mass transit. All are subsidized.