Why the US Sucks at Building Public Transit

American cities are facing a transportation crisis. There’s terrible traffic. Public transit doesn’t work or go where people need it to. The cities are growing, but newcomers are faced with the prospects of paying high rents for reasonable commutes or lower rents for dreary, frustrating daily treks. Nearly all Americans, including those in cities, face a dire choice: spend thousands of dollars a year owning a car and sitting in traffic, or sacrifice hours every day on ramshackle public transit getting where they need to go. Things are so broken that, increasingly, they do both. Nationwide, three out of every four commuters drive alone. The rate in metro areas is not much different.

“Without an integrated system of transit in our metropolitan areas the great anticipated growth will become a dream that will fail,” predicted Ralph Merritt, general manager of the Los Angeles Metropolitan Transportation Authority, “because people cannot move freely, safely, rapidly, and economically from where they live to where they work.”

Although Merritt’s words could just as well apply today, he said them 66 years ago in 1954. This is a crisis facing American cities right now in 2020, but it’s an old crisis. The only thing that has changed is the problem has gotten worse.

Like most crises, there is no single cause. Our cities, and our federal government, have made a lot of mistakes. Some were obvious at the time, others only in hindsight, but most have been a combination of the two. We keep doing things that stopped being good ideas a long time ago.

Many of those mistakes have to do with housing policy, which is inextricably linked to transportation policy. But the most obvious cause of our transportation crisis is a simple one: America sucks at building public transportation.

Why is this? Why does the U.S. suck at building good, useful public transit?

It’s a question that has vexed me for years. Just when I think I’ve figured it out, some other facet I had never previously considered comes to my attention. I have spoken to a dozen transit experts and historians. I have read several histories of American mass transportation policy written by independent scholars as well as government agencies. I've scoured federal archives and interviewed employees of transit agencies planning their own big projects. I’ve analyzed budgets and construction costs and compared them to our international peers. The tangle of American governmental dysfunction is so profound, digging into it can feel like undoing a rubber band ball with your teeth.

But the failure itself is simple and obvious. It’s apparent to anyone who has traveled abroad in the last several decades. Whether it’s traditional subway and commuter rail systems, modern streetcars and light rails, high-speed intercity rail, or even the humble bus with dedicated lanes and train-like stops, the U.S. lags perilously behind. It is a national embarrassment and a major reason our cities are less pleasant, more expensive places to live.

Just to name a few recent accomplishments abroad lacking an American parallel: Paris has the Grand Paris Express, 120 miles and 68 stations of new lines, plus a host of new trams and express bus lines with dedicated lanes. Moscow is building 98 new miles and 79 new stations for its Metro. At two years delayed and three billion pounds over budget, London’s Crossrail qualifies as a scandalous by European standards. But when it opens—perhaps in 2021—it will provide 73 miles of new rapid, frequent trains across greater London, including right through the center of the city. Since the 1990s, Madrid’s Metro has added more than 100 kilometers to its system. There are numerous examples of highly functioning and useful public transit systems in Latin America, which also invented the Bus Rapid Transit, a hybrid system with enclosed stops and dedicated lanes. China, which had basically no rapid transit through 1990, now has 25 cities with comprehensive rail systems, including seven of the world’s 12 largest metro networks by length.

Of course, China’s massive central government means it can build what it wants when and where it wants. But it’s hardly just China and other authoritarian regimes embarrassing the U.S. when it comes to transit construction. Consider, for example, high speed rail, or trains between cities capable of going faster than 120 mph. Over the last 30 years, almost two dozen countries have built true high-speed rail networks, according to transportation expert Yonah Freemark. The U.S. has a grand total of 34 miles of high-speed track.

This isn’t to say the U.S. has built nothing in the same time period. Freemark, one of the most thorough chroniclers of American transportation projects, calculated that the U.S. spent more than $47 billion on 1,200 miles of new and expanded transit lines in the decade from 2010 to 2019 (most of that mileage has been on bus routes).

That may sound like a lot, and at first glance it can seem like the U.S. has made some progress. There are now 93 miles of light rail in Dallas, 60 miles in Portland, and 87.5 combined light and commuter rail miles in Denver. Los Angeles, Seattle, Houston, San Diego, Sacramento, Phoenix, and others can cite similar improvements. For all their flaws, these are transit systems that didn’t exist 40 years ago.

But these systems, with the possible exception of Portland’s, do have one thing in common: they’re not especially useful because they’re not big enough and don’t go where people need them to. There is no perfect metric to evaluate the usefulness of a transit system, but, the most obvious failure is these systems haven’t changed their cities. Few people rely on them. As a general rule, these light rail systems serve fewer than 30 million passenger trips a year (LA has more, although as a percentage of the metro area population its usage is in line with other new systems). Even in cities of millions of people like Houston and Phoenix, light rail systems serve fewer than half that. Meanwhile, the Grand Paris Express and Crossrail are projecting ridership in the millions per day.

The basic truth is nearly everyone still depends on their cars even in cities with soul-destroying traffic. By any definition, the last half-century of American transportation policy has been a dismal failure.

But the problem isn’t limited just to new systems with growing pains. Older American cities with legacy systems have barely expanded to meet the growing footprint of their metro areas, as London and Paris are. The subway maps of New York, Boston, Chicago, and Philadelphia look almost identical as they did in 1950; in some cases, they’ve actually shrunk.

Simply put, the U.S. builds less public transit per urban dweller than its peer countries. Freemark found U.S. cities “added an average of fewer than 2 miles of urban bus improvements per million inhabitants—and fewer than 1 mile of rail improvements.” Meanwhile, France added 10 miles of buses and 3 miles of rail per million inhabitants in that same time period.

There is, of course, no simple answer why our transportation systems are broken, in much the same way there’s no simple answer to why our healthcare system is broken or why our criminal justice system is broken, beyond, as Freemark put it, that our “dysfunctional, irascible political system [is] woefully unprepared to commit to anything particularly significant.”

Ultimately, this is not about trains and buses. This is about a political system uninterested in reform, a system unconcerned with fixing what’s broken. If we can understand how politics failed American transportation systems, perhaps we can make the solution part of broader reform that must occur if American government is to start addressing the needs of the people in all aspects of life, from health care to criminal justice to housing to employment law to digital privacy to climate change.

It’s more important to understand all those causes now than ever. Building lots of public transit fast is, according to the Department of Transportation, a key front in the fight against climate change, because transportation accounts for about 30 percent of U.S. emissions, most of that from private automobiles. Are we up for the task? Can we, as a nation, build the infrastructure we desperately need to create a more sustainable world?

The answer to that question depends on understanding why we have failed so miserably up to this point. While researching the question of why our public transit is so bad, I’ve encountered a series of partial but ultimately incomplete explanations. If you don’t feel like descending into the transit nerd tunnel with me, here’s the tl;dr version:

• Everything costs too much
• We build highways instead
• We don't plan well
• People don't trust the government to build things so they vote against projects under the assumption they will be executed poorly and waste taxpayer dollars
• We don't give transit agencies enough money to run good service which erodes political support to have more of it
• There are too many agencies at all levels of government, especially at the local level, and not enough coordination between them
• Our newer cities are sprawled out which makes good transit hard, and our older cities are too paralyzed by political dysfunction to expand the systems they have
• As a result of generations of privatization efforts by all levels of government, in the rare event we do actually get to build stuff there is not enough expertise within the agencies to do it well

The good news is all of this is fixable. At least, that’s what Freemark believes. “The idea that we can’t build new systems is ridiculous,” he told me in an interview. “We just have to assemble the political interest and excitement to make those things happen.”

Before we go any further, it’s important to dispel a pernicious myth that has perpetuated in the United States about public transportation. This is the idea that transit ought to pay for itself just like any other business.

This was a popular position in local, state, and federal governments until the mid-20th century. It is also the founding principle of public authorities, like the Metropolitan Transportation Authority that oversees much of greater New York’s transit, which are legally required to balance their budgets every year. The concept is that well-run public transit ought to be profitable.

The problem—well, just one of the problems—with this philosophy is it’s based on a totally fictitious belief that the New York City subway once was a good business, or that the Boston subway once was a self-sustaining operation.

This was never true. “There was always a subsidy somewhere,” Jeff Davis, senior fellow at the think tank Eno Transportation Center said. Streetcars and early subways were paid for by wealthy financiers, real estate speculators, and electric companies, among others. The speculators bought cheap land on the outskirts of town and then built transportation that went there before selling the land for a tidy profit. Back in the day when lights were the main use of electricity, electric companies faced a huge surge at night. Streetcars were a convenient use of that excess electric capacity during the day when demand was lower. And, as the 19th century became the 20th, financiers (mistakenly) thought rapid transit would be a great investment, typically as part of an arrangement we now commonly refer to as public-private partnerships that required transit companies to keep fares low, usually at five cents.

Then it all slowly fell apart. Inflation jacked up costs, but transit companies were legally obligated to keep fares the same per their agreements with cities. The Great Depression hit. Real estate speculators sold off all their land and no longer cared about the transit connections. The public utility companies were forced to sell off their streetcar stakes by Congress under an antitrust provision. Long-term maintenance and upkeep rendered short-term profits illusory. Although most commuters still used transit through the 1940s, people tended to use private automobiles for recreational trips. Bills for decades of deferred maintenance came due. Streetcars went bankrupt. Local governments picked up the slack, and as part of the transition, closed the electric streetcars and converted those routes to buses. By the 1960s, most every transit system had either closed down or was under the auspices of some level of local government.

“And then that subsidy became an explicit job of the local government to subsidize and take over management,” Davis said. Private subsidies were replaced by public ones, just at the time when government was deeply, fundamentally uninterested in public transit. Because in the mid-20th century, cars were the future.

From 1950 to 2017, the U.S. constructed 871,496 miles of roads, enough to go to the Moon, come back, return to the Moon again, and then get two-thirds of the way back to Earth. The pace has slowed in the last few decades, but barely. Thirty-seven percent of those miles have been built since 1985.

As traffic increased, it was accepted policy to widen a lot of roads under the mistaken belief this would reduce traffic. The Federal Highway Administration only started tracking lane-miles built in 1980, but in the 37 years between then and 2017 we added 881,918 lane-miles to our some 8.8 million lane-miles of road, an 11 percent increase. Urban areas in particular added 30,511 new lane-miles to freeways since 1993, an increase of 42 percent, according to the non-profit Transportation for America, which went on to call this program of building more lanes in a misguided attempt to reduce traffic a “congestion con.”

In the meantime, the U.S. barely built any new rail. The Bureau of Transportation Statistics only started tracking rail miles in 1985, but from that year through 2017 the U.S. constructed 6,247 miles of commuter rail, heavy rail, and light rail combined. That’s only 195 miles a year on average, compared to 10,017 miles of roads per year during that same time. In fact, the pace of building new transit has been so languid, America’s 20 largest metro areas have the same or even fewer miles of transit service (including bus routes) per capita than they did in 2003.

“It’s all about priorities,” said Jeff Brown, an urban planning professor at Florida State University. “What are the spending priorities that we’ve established?”

Of course, the short term cost of building a mile of road is lower than building a mile of transit, but that can be deceptive. According to Transportation for America, it costs $24,000 per lane-mile per year to maintain a road in good repair, and much more for those in disrepair, as many of America’s roads are. And that’s even before accounting for the strain on public services by encouraging and supporting sprawl where every mile of sewer, water, and power line serves fewer taxpayers.

Nevertheless, we’ve also spent much less money overall on transit compared to roads. These funding mechanisms are extremely confusing and have changed over time, but what has not changed is that roads always get a lot more.

Congress gives states roughly $40 billion a year for roads, according to Transportation For America, which can be spent either on new roads or maintenance at the states' discretion. Meanwhile, public transit agencies have to compete for only $2.3 billion in annual transit funding for big projects such as extending rail lines or building new ones, some $37.7 billion less than what states get for roads (the feds dole out an additional $7.5 billion a year for maintenance and buying new subway cars and buses).

That $40 billion a year in road money is given out to states based on a formula. It’s automatic, and states can spend that money however they wish. Not so with transit money. Transit agencies have to apply for funding for individual projects.

And should the transit agency’s project be deemed worthy of federal funding, the federal government will subsidize a much smaller percentage of the project costs than it will for roads. Transit agencies can get a maximum of 50 percent of the project cost covered by the feds, whereas roads can get up to 80 percent (down from 90 percent during the highway spending spree of the 20th Century).

And this is just at the federal level. The discrepancy between road and transit funding is even wider at the state level, Freemark says, where legislatures are typically dominated by rural interests.

Brown, the Florida State professor, said the numbers don’t lie. “It’s not a sufficient amount of money to support grand project ideas.”

Of course, many people believe it is not the federal government’s role to be paying for mass transportation because it’s a local issue (rarely is a similar argument made about roads). This was very much up for debate when the Urban Mass Transportation Administration (now a part of the Federal Transit Administration) was created in 1964. The upshot was that there’s no clear reason why the federal government should be subsidizing road construction, home mortgages, auto fuel, and any number of other things but not mass transportation. Plus, in light of the local and state government failures to pay for transit, if not the federal government, then who? Tellingly, the UMTA was founded under the Welfare Clause of the Constitution, not the Commerce Clause that authorized highway construction, because it is good for cities to have good transportation.

From nearly any vantage point, this road-heavy, transit-lite approach has been a disaster for American cities. We’ve spent hundreds of billions of dollars constructing and maintaining an unsustainable roadway network, and traffic has only gotten worse to boot. In 2015, California’s Department of Transportation, which supervised some of the most fervent highway construction in the nation during the 20th century, came right out and admitted this didn’t work. More roads means more traffic. So, the state is no longer going to keep widening roads to relieve congestion.

Not only is there not enough money to go around, but it has to be shared by all the states. Federal rules require that no single project gets too big a slice.

“You can’t ask for so much money in a single year as to crowd out everyone else,” explained Davis. For example, he said, the Federal Transit Administration (FTA) under the Obama administration told Los Angeles that it wouldn’t get all the money it wanted for the Westside Purple Line as one big extension through Beverly Hills and into Westwood. So, LA broke it up into three segments, with construction on Phase I beginning in 2014. Each got its own cost-benefit analysis, planning, and studies, and waited a few years between applications, which drives up costs. In February, LA received its grant for the third and final segment, a $1.3 billion payout that will cover just 36 percent of the cost. It is expected to be completed in 2027, meaning it will take 13 years to build a nine mile extension.

“Usually congressional and even executive branch political realities mean they spread the peanut butter around,” said Sarah Jo Peterson, author of Planning the Home Front, “and, when there isn't much peanut butter, they spread it thinly.”

Not only does the money get spread too thinly, but once cities do get their money, they waste a lot of it.

“In the cities where rail transit works best,” Davis observed, “costs have just gotten out of control.” This is especially true for megaprojects, huge public works that cost billions of dollars.

I could spend an entire article on this subject alone and not even scratch the surface of just how profoundly screwed American megaproject costs are. Indeed, many writers and researchers have done exactly that, and one researcher in particular, Alon Levy, has more or less made a name for themselves on this subject.

New York City is responsible for the most expensive mile of subway track on Earth, at $3.5 billion per mile, the first segment of the Second Avenue Subway. The second phase is projected to crush that record. The Metropolitan Transportation Authority, which runs the subway and commuter rail system, is also some two decades late and $8 billion over budget on the $11 billion East Side Access project, which will bring Long Island Railroad trains into Grand Central, a 15 minute walk from Penn Station where Long Island Railroad trains currently go.

The problem is hardly limited to New York. California’s high speed rail project has given new definition to the term “boondoggle.” And, as Levy has documented, San Francisco, Los Angeles, Seattle, Boston, and D.C., among others, all build subways and light rail lines at much higher costs than European cities.

“Nearly all American urban rail projects cost much more than their European counterparts do,” Levy wrote in Citylab. “The cheaper ones cost twice as much, and the more expensive ones about seven times as much.” This includes both heavy rail (subways) and light rail. “Only a handful of American [light rail] lines come in cheaper than $100 million per mile, the upper limit for French light rail.”

There are a lot of reasons for this, including:

• Over-engineered stations
• Arcane labor rules that inhibit productivity such as requiring more employees to work at a machine than is necessary
• A lack of cooperation between agencies

But cost overruns are not a new problem for American transit.

“Cost escalation has been going on for some time,” Davis of Eno Center said. The D.C. Metro was initially slated to cost $2.5 billion but ended up with a $10 billion bill. Dating back to the Ford administration, the federal government started changing grant rules so the feds wouldn’t be on the hook for the inevitable cost overruns, leaving any eventualities to the transit agency building it and the local government overseeing it.

Some cost overruns are attributable to unforeseen circumstances, but across the board, we are very bad at estimating how much these projects will cost to begin with. Sometimes, agencies give low estimates in order to make projects more politically palatable, knowing a realistic assessment will get shot down. D.C.’s initial estimate of $2.5 billion, according to George Mason University historian and author of The Great Society Subway: A history of the Washington Metro Zachary Schrag, was “never terribly realistic.”

Depressingly, it seems we gave up on ever building necessary infrastructure for the same price as other countries decades ago. Schrag also quoted Jim Caywood, head of the engineering firm that helped design Metro, as saying “there’s no way in this world that you can build a mammoth public works project such as Metro within a reasonable budget with all the outside influences. They won’t let you do it.”

What are those “outside influences” Caywood referenced? The big one is politics.

“Transportation planning is not just a matter of letting the engineers find the best solution to a technical problem,” Schrag wrote, “but a political process in which competing priorities must be resolved by negotiation among interest groups.”

If there’s one point on which all the experts I spoke to agree the most, it is that transportation is politics. A 1988 deep dive into the construction of the Bay Area Rapid Transit (BART) system by Portland State University professor Sy Adler found any vague proposal for a transit project, whether it be highway or rail, produces competing coalitions with their own self-interests. Maybe they want to spur development in their own downtown area or make it easy for commuters to live in their suburb. These factions then weaponize the options on the table for their preferred ends. The protracted debates result in entire regions losing focus over why they wanted to build a new transit system in the first place. Over time, it becomes a battle not of which option solves a given problem, but re-defining what the problem is.

Meanwhile, another cohort of interest groups form to stop projects they don’t want. Typically, these are neighborhood associations that don’t want a transit line coming through their block, either out of fear of construction impacts or racist concerns that it’ll disrupt the segregation of their urban area. Sometimes, they are not just neighborhood groups but entire regions.

In the 1970s, in what was later called “referendums on race,” Atlanta’s suburban and overwhelmingly white Cobb, Clayton, and Gwinnett counties voted out of the region’s MARTA system before it was even built. In 1988, five years after voting in favor of a 200-mile light rail system with local financing, Dallas voters refused to approve bonds to pay for construction, growing skittish on the whole prospect due to the oil crisis which hurt the local economy, according to Indiana University professor George Smerk’s history of the government’s mass transportation policy. Seattle, Detroit, and scores of other cities either voted against major transportation projects or approved watered down versions of original plans thanks to local opposition.

What is undoubtedly clear is every transit project is first and foremost a political project, and political projects are about consensus-building. This gets us not the projects we need but the projects we deserve.

To take just one of scores of possible examples because it affected me personally: back in the 1970s the University of Maryland rejected plans to have the D.C. Metro’s Green Line stop on campus, again for predominantly racist reasons. This forced “a complicated redesign” that “later caused commotion in College Park,” Schrag wrote. Today, anyone looking to take public transportation to the university, with 41,000 students and 14,000 faculty and staff, must either take a shuttle bus from campus or walk at least 20 minutes each way. Repeat these fights dozens of times per project and it’s no longer so difficult to envision how they end up getting relegated to land the public already owns regardless of how useful it is like freeway medians or don’t get built at all.

Meanwhile, a great political shift occurred in the United States that made transit’s prospects even worse. First was the Reagan-era movement away from services provided by the government and towards private enterprise.

Transit was not spared. When Miami’s Metrorail opened, Reagan derided the “$1 billion federal subsidy” that “serves less than 10,000 daily riders” as a prime example of government waste. Better for the government to have bought everyone a limousine, Reagan quipped.

Nevermind that all of those numbers were incorrect and deeply misleading because the project hadn’t even been completed yet, according to the Sun-Sentinel. But factual errors aside, there was a larger ideological one. “Even if Metrorail doesn't turn a profit,” the paper said, “it will be performing a valuable service. Without it, thousands of new commuters would be forced back into their cars, making the roads even more overcrowded.”

Ironically, some of the blame for the wastefulness of federal transit money belongs to Reagan himself. He spent considerable effort trying to kill the main transit grant program, according to Davis, but Congress wouldn’t let him because these projects were often popular.

In order to keep the funding going, Congress had to resort to doling out the money through annual appropriations—in other words, the 435 members of the House of Representatives, with all its byzantine committees and rules, deciding for itself which projects to fund rather than career experts in the Federal Transit Administration—through a process called earmarking. In this way, transit projects became just another horse to trade.

“The nature of earmarking is that since there are 435 House districts and 50 states, and only so much money to go around, things get split more widely than they would if the Administration just got to pick a few.” Davis continued: “Ted Stevens [longtime Alaska Senator and chair of the Senate Appropriations Committee from 1997 to 2005] used to put a million dollar grants to little towns all over Alaska. I don't know what the hell they were for.”

Not only did it become fashionable to slash funds for big transit projects, but so too was it the sign of the times to slash agency budgets as well. Expertise then migrated to the private sector, in many cases to the very consultants and engineering firms hired to execute the few projects that got done. As a result, agencies were—and remain—ill-equipped to make big decisions on big projects, who hire those aforementioned consultants, who in turn charge a pretty penny for their services.

“We often don't have really expert public staff making decisions, making some key decisions at least,” said Eric Eidlin, a former FTA planner. “We've given over that responsibility to consultants that have a profit motive. I don't mean to say that the consultants have this desire to subvert the public interest or anything, it's just not their job, right?”

Karen Trapenberg Frick, an urban planner at UC Berkeley who used to work for the Bay Area’s planning commission, echoed Eidlin’s point and said it had a real impact on what agencies were able to do, only further undermining the public’s willingness to give them money for big projects.

“There are certain cities where when I was a planner a long time ago and now, it's the same complaint: we give the city money but they can't move the project through because they don't have the staff to do it,” she explained. “And we don't have the staff to do it because there's been this whole neoliberal mind shift that the public sector can't do a good job.”

There’s a very sad irony here. The Reagan era cuts were ostensibly designed to make the public sector more efficient by harnessing the power of the market, but instead it made public agencies reliant on for-profit contractors that jack up costs, only making government less efficient and more wasteful.

“When there's no in-house public sector expertise, the ability to deliver projects quickly or efficiently is compromised,” Eidlin said. “And time is money, too.”

So far, I’ve focused on the federal side of things because it has a lot of money and power. But what level of government is the right one to make decisions about massive transportation projects? Although there is no obvious right answer, it feels like the U.S. has discovered an awful lot of wrong ones.

As a nation, local authority is our founding principle. We fought a revolution to achieve it, wrote a (bad) set of rules to maintain it, scrapped those, then wrote a new set of rules we have been arguing about ever since. Most of those arguments have been about whether the federal or state governments should determine how we live.

But unprecedented depressions and world wars have a funny way of harnessing big government power, and the feds continued to flex those newly-discovered muscles as American cities deteriorated in the years afterward. From New York to Los Angeles and in dozens of cities in between, so-called “urban renewal” programs used federal dollars to quite literally tear down and rebuild massive sections of cities from scratch, sometimes in order to build a highway through the demolished portion. One of the many legacies of this program, which destroyed entire neighborhoods, was a growing distrust in the government to sensitively execute centrally planned projects. The preferred remedy was to have more local control, neighborhood by neighborhood.

This approach has its merits, but for transportation it has serious drawbacks. Whether they be subways, light rail, bus routes, or even the humble bike lane, any transportation worth using is a network that allows people to get from one side of a city to another quickly and efficiently. Giving substantial input or even veto power to individual communities along that network undermines the entire concept.

“Transit is fundamentally regional,” Eidlin said, “And I really feel like our general population and our decision makers don't universally agree with that or even had that epiphany yet.”

Just as too much hyperlocal control can stymie useful transit, putting transit under the auspices of entire states can have downsides, too. Several of the country’s biggest transit systems including New York, Boston, and Washington, D.C. are controlled not by local authorities but state (or in D.C.’s case, quasi-federal) bodies. This means taxpayers who don’t obviously benefit from the system pay into it, a constant source of political tension. And when proposed projects cross state lines, it opens up a prolonged debate about who pays for what share, a fight that often takes years or decades to resolve.

Put the lack of funds for transit together with our country’s general desire to give local control as close to the individual citizen level as possible, and we’re left with a contradictory system where every limb and appendage fights the others. The lack of funds dedicated to transit means higher and higher levels of government—the ones with more and more money—often have control over transit, either by law or by practice. But those same agencies must seek local consensus for what are not local projects, a time-consuming and expensive proposition at best and a poison pill at worst.

This desire for local control yields bizarre outcomes. For example, Eidlin is working on a transportation hub project in San Jose, CA. Four different public agencies are involved, each for a different jurisdiction that will meet at the hub (this is indicative of the Bay Area, which has 27 transit operators and 151—yes, 151—transit agencies). As a result, Eidlin says much of the project’s work at this stage is not on the project itself, but administrative tasks to keep all the agencies up to speed.

“We value local control so much and we fund so many things locally that we never stop and ask,” Eidlin said, “what's the right level of government at which to be addressing a public issue?”

As dire as the American transit landscape is, there are specs of hope. Federal funds are no longer given out through earmarks; that stopped in 2010. Now, the FTA grades projects based on merit. And some metro areas have big plans. Los Angeles and Seattle voters have opted to raise their sales taxes slightly to fund tens of billions of dollars in transit upgrades that could significantly improve their region.

But we need much bigger solutions, not only to build transit systems faster and more efficiently, but to run them better, too. In the vicious cycle of transit funding, agencies that are perceived as wasteful or bad at providing services have a harder time getting money from politicians, which then makes it harder to run a good transit service. This cycle must be broken.

More money for transit would obviously help. Bernie Sanders has proposed $300 billion for public transit by 2030 and $607 billion for a high speed rail network (Joe Biden, in an excellent distillation of the failures of American transportation policy to date, does not commit any dollar amounts to these issues in his platform, but does commit $50 billion in his first year to repair roads, highways, and bridges). That would be a lot more money where it’s desperately needed, and polling suggests it’s a popular platform with majority support.

The most noteworthy part of Sanders' platform, however, is not the money. It’s the framework under which it is proposed: the Green New Deal.

This, says Florida State University’s Jeff Brown, fits with the history of how big transit projects are proposed. “Transit, in most places, has very much been an afterthought or a reaction to some other perceived crisis,” he said. Traditionally, that crisis has been traffic. For periods in the 1970s and 1980s, it was the oil crises. Sanders, however, clearly puts better public transportation within the framework of the climate crisis.

But the very concept of tying transit construction to a crisis misses the point. Transit does address those issues, but it is more than that. We will never build good transit until we jettison the century-old misconception that it is a business the government happens to run out of necessity. Rather, public transportation is a public good on its own merits, good times and bad. Allowing people to move about their cities cheaply, efficiently, and quickly makes cities more productive and better places to live and has numerous knock-on public health, environmental, social, and economic effects. Public transit funding ought not to be a response to any crisis. It ought to be as natural a government service as trash collection.

On the other hand, framing transit as a fight against traffic is a losing battle, because it doesn’t take very many cars to create traffic. It is, as transit planner Jarrett Walker argues, a matter of geometry. It will always appear to a certain type of person that money was wasted. But positing that transit is a way for city dwellers to live better, more pleasant lives is a winning platform, as politicians across Europe can attest.

We also have to work out what the right level of government is to make transit decisions. New York, D.C., and San Francisco in particular have complicated and bizarre governance structures for their transit agencies. Most of these structures were created in mid-century when good governance types replicated the corporate boardroom as the ideal of good governance. History has proven this approach hopelessly naive. Transit is politics. It’s time to, as Freemark has argued, put transit squarely within the responsibility of one elected official who is clearly accountable.

None of this solves what may be the biggest impediment to good American public transit: costs. The solutions here are not easy. Hell, as Josh Barro of New York has pointed out, and I've also learned, we don’t even fully understand the problem. At the very least, fixing it requires cultivating long-term expertise on the local level so agencies aren’t reinventing the wheel the rare occasions they’re given enough money to undertake megaprojects. It also might require, as Laura Tolkoff of the San Francisco-based non-profit SPUR suggested, establishing governmental entities with in-house megaproject expertise, weaning the transit world off relying on expensive contractors and consultants and onto agencies looking out for the taxpayers’ interest, not the stock market’s.

These are just a handful of the high-level suggestions I learned while reporting this story. I will keep reporting on this and learning more, and you should contact me if you work in a transit-related industry and know anything I ought to know. But one thing we must always keep in mind is the answers are out there.

“[The U.S.] needs to learn what works in Japan, France, Germany, Switzerland, Sweden, the Netherlands, Denmark, South Korea, Spain, Italy, Singapore, Belgium, Norway, Taiwan, Finland, Austria,” Levy wrote. “It needs to learn how to plan around cooperation between different agencies and operators, how to integrate infrastructure and technology, how to use 21st-century engineering.”

To that end, Levy and fellow researcher Eric Goldwyn just received a two-year grant from the John Arnold Foundation via New York University to study why U.S. construction costs are so high. And they’re looking to hire a research associate, preferably one with language skills other than English. “We are particularly looking to extend our coverage outside countries where information is readily available in English,” their job posting for the project says.

“Imitate,” Levy advises. “Don’t innovate.”

Correction: A previous version of this article stated "we added 881,918 lane-miles to our some four million lane-miles of road, an 11 percent increase." The U.S. had 8.8 million lane-miles of roads as of 2017, not four million.