Randal O’Toole’s recent book, Romance of the Rails,1 is a slam-dunk. Actually, that is an understatement. The book is full of slam-dunks. In chapter after chapter, O’Toole, a long-time fan of railroads, puts his fandom aside and shows what a disaster government subsidies to, and regulations of, rail transportation have been. The book, subtitled “Why the Passenger Trains We Love Are Not the Transportation We Need,” is a fact-filled and numerate critique of some aspects of government-subsidized rail transit, whether it be passenger trains, subways, low-capacity rail, or high-speed rail. His bottom line is that in case after case, governments have turned the clock back on transportation, resisting and trying to reverse its progress.

O’Toole, a senior fellow at the Cato Institute, has been writing about transportation for many years and the depth of his knowledge shows. Chapters One through Five are largely about the successes of the early railroads. Chapters Six and Seven document the decline of urban rail transit and intercity passenger trains in response to market forces. Chapters Eight through Seventeen document the hugely expensive methods governments have used to preserve and extend rail transit, including subways, with little or no positive effect on passenger use of those modes of transit.

One of the most interesting and positive characters in the book is Ralph Budd, an early innovator in rail transportation. In 1934, Budd delivered the Zephyr to the Burlington Railroad. The Zephyr was a “streamliner,” a light train that could travel at over 100 miles an hour using minimal fuel. To show off its virtues on a 1,015-mile trip from Denver to Chicago, Burlington had guards at all of its 1,700 grade crossings so that the train would not have to slow. The train arrived just 13 hours and five minutes later. The cost of the fuel used? $15. I emailed O’Toole to ask if that was a typo. It was not. O’Toole explained that the Zephyr used only 412 gallons of diesel fuel, one of the cheapest fuels around. As a fellow train buff, I recall seeing a mid-1930s ad in Fortune with the saying “Have dinner in Chicago and breakfast in Denver.”

The rest of the book is pretty much downhill from there—not, I hasten to add, in quality, but in negative finding after negative finding.

One of the first modes of rail travel to face a long-term decline was streetcars. Streetcar route-miles peaked in 1919, a century ago. And streetcar trips fell along with route-miles. There were two main causes: cars and buses. Both had the advantage that they were not on rails. Cars could take their passengers wherever they wanted to go and buses could change their routes in response to changes in demand. Not surprisingly, after two decades of decline, there was a boom in streetcar trips during World War II. O’Toole mentions the cause—rationing—but does not specify what was rationed. Gasoline was stringently rationed, and the federal government banned the production of cars for private use in early 1942.

No book that covers the decline of urban rail transit would be complete without the obligatory section on the alleged streetcar conspiracy. I will not tell all the details here. The highlights are that, in 1974, an antitrust lawyer named Bradford Snell claimed in front of a Congressional committee that General Motors had tried to destroy urban rail systems in order to sell more cars. Snell averred that, in 1947, the federal government had brought an antitrust suit against General Motors for trying to destroy transit systems. A simple search of 1947 antitrust cases by a Congressional aide would have shown Snell’s statement to be false: the 1947 antitrust charge against GM was for monopolizing the sale of its buses. Moreover, although O’Toole does not point this out, basic economics says that monopolizing an industry causes the price of the industry’s output to rise. That makes Snell’s charge against GM all the more ludicrous because a higher price for buses makes the switch from rail less likely, not more. I recall George Hilton, my transportation economics professor at UCLA, telling us in class how absurd Snell’s claims were. O’Toole references Hilton’s testimony on the issue before Congress. O’Toole also quotes two transit historians’ conclusion that if there was a conspiracy to destroy streetcar companies, the feds should “indict everyone who bought an automobile” between 1920 and 1950.

Unfortunately, the myth did not die. It lived on in popular books and in the 1988 Disney movie Who Framed Roger Rabbit. O’Toole even cites Portland State University planning Professor Martha Bianco’s argument that the movie and an error-filled 60 Minutes episode helped persuade Congress to pass the Clean Air Act of 1990 and the Intermodal Surface Transportation Efficiency Act (ISTEA) of 1991, both of which penalized cars and started large subsidies to rail transit.

“O’Toole argues that the main causes of the decline in passenger rides were interstate highways, air travel, and a 1947 ICC regulation that set speed limits for trains.”

O’Toole argues that the main causes of the decline in passenger rides were interstate highways, air travel, and a 1947 Interstate Commerce Commission (ICC) regulation that set speed limits for trains. In 1946, in Naperville, Illinois, a fast-moving train had crashed into the train in front of it, killing 45 people. Burlington, which owned both trains, responded by changing its policies for high-speed trains. Previously, it had allowed them to run just three minutes apart. Henceforth, that was raised to 15 minutes. O’Toole points out that this one change would have prevented the Naperville crash. But the Interstate Commerce Commission was not satisfied. It insisted on a 79-mph speed limit unless the railroads installed equipment to display signals inside the locomotive cabs. The Pennsylvania Railroad estimated that the new regulation would cost about $4,000 per track mile and $2,260 per locomotive, for a total cost of $100 million or, notes O’Toole, more than $1 billion in today’s dollars. Many railroads balked at the costs and, instead, reduced their speeds. That hobbled them in their competition with cars and, increasingly, with airlines.

Throughout the 1950s and 1960s, the demand for passenger train travel had declined so much that railroads dropped passenger routes when they could get the ICC’s permission to do so. The Hosmer report, a 1958 ICC study, predicted that sleeping car service would disappear by 1965 and coach service by 1970. On the demand side, as noted above, cars and airplanes were eating into the passenger railroads’ market share. Also, some studies at the time concluded that passenger train travel was what economists call an inferior good: when people’s income rose by one percent, train travel fell by 0.6 percent. On the cost side, the railroads were hobbled by federal union work rules established in 1919. These rules required railroads to pay a full day’s pay to engine crews for every 100 miles they worked and train crews a full day’s pay for every 150 miles. When the rules were established, notes O’Toole, train speeds averaged about 20 miles per hour. By 1958, they averaged 40 mph. In 1970, Congress passed, and President Nixon signed, a law that created the National Railroad Passenger Corporation, subsequently known as Amtrak. The earlier-mentioned George Hilton, a train aficionado who, like O’Toole, never let his fandom color his analysis, wrote, “Amtrak seems to me to have all the antique charm of a Travelodge and a Denny’s—and to be equally worthy of enthusiasm.”

Among those who should not be enthusiastic are taxpayers. O’Toole quotes Anthony Haswell, who, he writes, is “considered the father of Amtrak,” on what a bad deal Amtrak is. “Amtrak has cost the taxpayers a lot more than $20 billion,” wrote Haswell, but taxpayers “have gotten back a pretty skimpy return.” O’Toole dissects Amtrak’s 2016 press release on its financial statement and shows that Amtrak misled by counting subsidies from state governments as “passenger revenues” and not even mentioning the large depreciation of Amtrak’s capital. Correcting for those two items, O’Toole shows that Amtrak’s reported small loss of $227 million for 2016 was, in reality, more than five times as large, at $1,128 million. Even the vaunted Northeast Corridor, he notes, costs taxpayers about 11 cents per passenger mile.

For more on these topics, see Surface Freight Transportation Deregulation, by Thomas Gale Moore and Urban Transportation, by Kenneth A. Small in the Concise Encyclopedia of Economics.. See also Bent Flyvbjerg on Megaprojects, by Bent Flyvbjerg on EconTalk

Buses, by contrast, have filled the city-to-city niche for trips of a few hundred miles. So-called Megabuses pick up passengers at designated curbsides, have few or no intermediate stops between major cities, have comfortable seats, and have Wi-Fi and power ports at each seat. A quick check of one company, Boltbus, shows a travel time of 4 to 4.5 hours between Washington, D.C. and New York City, along with a fare of $28 to $31. O’Toole notes that intercity buses carry twice as many passengers as Amtrak. The subsidy to buses per passenger mile, he reports—based on road construction and maintenance costs that are not paid out of fuel taxes and highway user fees—is about 1/8 of a penny, which is 1.1 percent of the Amtrak subsidy per passenger mile for the Northeast Corridor.

O’Toole devotes most of the remainder of the book to bus transit and rail transit within cities, including subways and “light rail.” The losses there are far greater than the losses for Amtrak, on the order of billions of dollars a year. One of the major pieces of legislation that caused losses was the Urban Mass Transit Act (UMTA) of 1964. Before it passed, writes O’Toole, “more than three out of four transit systems in the nation’s 100 largest cities” were privately owned, including systems in Philadelphia and Washington, D.C. A few transit systems, mainly those that were government-owned, lost money, but overall revenues exceeded operating costs by five percent. Ridership, however, was declining.

UMTA gave federal funds to city and state governments that took over transit companies. Can you guess what happened next? Local governments took over transit companies. Over the next twelve years, writes O’Toole, 70 major cities municipalized transit.

Once that happened, transit companies were no longer constrained by fare revenues because local governments would subsidize the transit companies that they now owned. Between 1964 and 1975, inflation-adjusted fare revenues fell by 18 percent, but operating costs grew by 46 percent. A major component of this increase in costs was an increase in worker pay and benefits. Transit unions, aware that local governments both wanted to avoid strikes and could draw on tax revenues, took advantage. From 1964 to 1975, average transit worker pay, adjusted for inflation, rose by almost 30 percent. Another major cause of the increase in operating costs was reduced worker productivity. Before UMTA, the private transit industry carried 55,000 to 60,000 passengers a year per operating employee. After municipalization, productivity fell to about 27,000 passengers per operating employee. Writes O’Toole: “This stands out as one of the biggest declines in employee productivity of any industry in history.” In 1975, Congress started to offer operating subsidies. By 2002, fares covered less than a third of operating costs and, since then, have been around 32 percent of operating costs.

In the late 1960s and early 1970s, local governments, partly to get federal funds, started subway systems in San Francisco (Bay Area Rapid Transit), Atlanta (Metropolitan Atlanta Regional Transit Authority), and Washington (Metro), where the odds of their being self-sufficient were close to zero. The cost of constructing the Bay Area Rapid Transit (BART) system was 50 percent greater than projected, which caused BART to ask the state legislature for the authority to collect a sales tax. The state legislature granted a 0.5 percentage point increase. Actual ridership turned out to be half the projected number, and operating costs were a whopping 375 percent more than projected. O’Toole notes that sales taxes are regressive, but BART riders are wealthier than average, and, quotes University of California, Berkeley planning Professor Melvin Webber, “Clearly, the poor are paying and the rich are riding.”

The experience with the Metropolitan Atlanta Regional Transit Authority (MARTA) was similar to that with BART. Government officials in both areas reacted similarly, cutting bus routes in order to shore up funding for the subways. For Atlanta, the result was that between 1985 and 2015, per capita transit ridership fell by two thirds. This was tragic because when MARTA was being debated, opponents proposed, instead, a rapid-bus system that would cost $53 million and would move as many riders as the proposed half-billion-dollar subway.

The Washington Metro experience resembles BART’s and MARTA’s—exaggerated projections of ridership and massive cost overruns. But Washington’s Metro is even worse. O’Toole gives chapter and verse on Metro’s legendary safety problems. One highlight: “By 2011, Metro was suffering from an average of nearly one cracked rail and several smoke-related incidents every week.” But not content to merely stay in the hole, Metro’s planners are digging deeper. Instead of using funds to upgrade its highly depreciated subway system, Metro is expanding by adding lines.

One of the most eye-popping and informative chapters of O’Toole’s book is titled “Low-Capacity Rail.” Low-capacity rail is what is usually called “light rail.” But light rail, notes O’Toole, is a misnomer. He writes, “A typical light-rail car built today weighs about 105,000 pounds, while a typical subway or heavy-rail car weighs 83,000 pounds.” Nor are the rails they ride on lighter than subway rails. Why, then, is it called light rail? O’Toole explains by drawing on the Glossary of Transit Terminology. It’s called “light” because it has a light volume traffic capacity. In short, light means low capacity. The real high capacity carriers, shows O’Toole, are buses. Not surprisingly, “light rail” does not clearly boost transit ridership. In ten of the 17 urban areas that have built “light rail” since 1980, trips per capita and transit’s share of commuting fell. Those two measures rose in only three of the 17 urban areas. The Los Angeles County transit agency’s experience is instructive. It cut bus service to minority neighborhoods to fund more-expensive rail lines to middle-class neighborhoods. The NAACP sued and got a court order restoring bus service for ten years. But after the court order expired, the LA transit agency cut bus service and built more rail lines. Result: the system lost five bus riders for every new light-rail rider. Interestingly, the fatality rate for light-rail riders is four times that of bus passengers.

The costs for light rail are eye-popping. Orlando’s SunRail, which opened in 2014, had only 1,824 daily roundtrip passengers in its first year of operation. In 2016, the local government agency running SunRail admitted that fare revenues were less than the cost of operating and maintaining the machines that sold tickets to riders. O’Toole estimates that Orlando could have saved money by giving a new Prius to every roundtrip rider every year.

Is high-speed rail for long trips a good idea? O’Toole shows that it’s not. He starts by showing how expensive high-speed rail has been, even in countries where its results have been touted. Consider Japan. Three private companies operate trains on the main island, Honshu, and make money. But all three companies that operate on the other islands lose money and are regularly subsidized. The lesson, writes O’Toole, is that high-speed rail can make sense in high-density corridors but not elsewhere. France has one high-speed train that makes money—between Paris and Lyon—and the rest lose money. Spain’s and Germany’s high-speed rail have done nothing to reduce car rides, but, instead, have cut into the market share of buses. The major change that has changed increased travel in Europe, notes O’Toole, is not high-speed trains but Europe’s 1992 airline deregulation, which led to low airfares. Possibly that lesson is getting across in California, where Governor Gavin Newsom, elected in 2018, after O’Toole’s book was published, announced in February that California’s proposed high-speed rail from Los Angeles to San Francisco, currently under construction, would be scaled back to a 171-mile railroad between the booming metropolises of Bakersfield (population 380,000) and Merced (population 81,743), with completion expected by 2027.

On his last page, O’Toole nicely sums up his message: “[N]ew transportation technologies have replaced trains and streetcars. Planes are faster and less expensive for long distances; cars and buses are more convenient for short distances; and there is no midrange distance at which passenger rail has an advantage over both cars and planes.”


[1] Randal O’Toole. Romance of the Rails: Why the Passenger Trains We Love Are Not the Transportation We Need. Cato Institute, 2018. (As an Amazon Associate, Econlib earns from qualifying purchases.)

*David R. Henderson is Emeritus Professor of Economics with the Graduate School of Business and Public Policy, Naval Postgraduate School in Monterey, California. He is also a Research Fellow with the Hoover Institution and a Senior Fellow with the Fraser Institute. He blogs at EconLog and is Editor of the Concise Encyclopedia of Economics.

For more articles by David R. Henderson, see the Archive.