Fixing America's Freeways

The private sector is reinventing our expressways, one lane at a time.


If you work anywhere in America outside of New York City, chances are you drive to work. That means you battle congestion twice every weekday. Rest assured that it's not your imagination: traffic is much worse than it used to be. 

The Texas Transportation Institute (TTI), which has been measuring the cost of traffic congestion in wasted time and fuel for three decades, estimates that in current dollars the traffic penalty rose from $24 billion in 1982 to $115 billion in 2009 (the latest year for which complete data are available). The average urban commuter wastes 34 hours a year in rush-hour congestion today, compared with just 14 hours in 1982.

Those average figures reflect all 439 urban areas. But if you live and work in one of the 20 largest metropolitan areas, you're stuck in an even worse jam. In Los Angeles, the average commuter wastes 63 hours a year due to congestion. In Dallas/Ft. Worth, it's 48 hours a year. And in the Washington, D.C., area, it's a whopping 70—almost nine full working days that drivers could have back if only freeways and streets delivered motorists at the advertised speed.

Congestion does far more harm than simply wasting time and fuel. By reducing the area you can traverse in, say, 30 minutes, congestion shrinks your "opportunity circle" of jobs, entertainment, housing, and even dating. Economists find that reduced opportunity circles due to congestion inhibit the best matches between skilled workers and employers, reducing the economic productivity of congested urban areas. The chief economist at the U.S. Department of Transportation has estimated that the true annual economic cost of congestion is at least double the $115 billion figure noted above.

Why does congestion keep getting worse, and what can be done about it? While there is no single answer to either question, a principal reason for ever worse congestion is that the demand for road space (especially on urban freeways) greatly exceeds the supply. That's because after the initial burst of freeway building in the 1960s and '70s, additions to freeways slowed way down while population and economic growth continued apace. Some analysts liken the result to cramming 10 pounds of potatoes into a five-pound sack.

From 1984 to 2009, the TTI estimates, daily traffic on the freeway systems of America's 18 most congested metro areas increased by 142 percent while their capacity (measured in lane-miles) grew only half as much (72 percent). In most cases, that means freeways today are attempting to handle vastly more traffic than they were designed for. That also explains why "rush hour" in major metro areas today runs about three hours in the morning and another three or four hours in the evening.

Again, those are averages. A few metro areas, such as Houston, have expanded freeways much more than 72 percent. But others, such as Los Angeles, San Francisco, San Diego, and Minneapolis, have added only 25 percent to 30 percent more capacity. TTI's 2010 Urban Mobility Report cites a strong inverse correlation between capacity and congestion: In the 13 metro areas that have expanded freeway capacity nearly enough to keep pace with travel demand, congestion is only about one-third worse today than it was in 1982, while the 42 metro areas where the gap between demand and capacity was greatest ended up with congestion about 140 percent worse in 2010 than in 1982.

Building Our Way Out of Congestion

Given the enormous cost of congestion and its relentless increase during the last 25 years, why haven't more metro areas at least attempted to keep pace with the growth in traffic? There are two main reasons.

First, it has become far more costly to widen freeways and build new ones as urban areas have filled in with expensive development. The cost is political as well as economic. Unlike in the early days of freeway building, you can't just bulldoze neighborhoods wherever a new freeway would make transportation sense (on the whole, a positive development), and anything you do build must go through costly and time-consuming environmental reviews. 

Second, many transportation planners and politicians believe that expanding road capacity is futile. If you build more, they say, the new lanes will simply fill up with more cars, and within a few years you'll be back where you started. This is only partly true. 

Some research does suggest that if you add a small amount of capacity in a place like Los Angeles, where roadway supply and demand are massively out of whack, the latent demand for faster travel during rush hours will fill the new capacity in short order. Still, the long-term data from TTI demonstrate clearly that if a metro area can keep increasing capacity in step with demand, it will end up with far less congestion than places that don't.

Yet because the notion that "we can't build our way out of congestion" is widely accepted, the long-range transportation plans of most large metro areas these days are premised not on making it easier for commuters to engage in their demonstrated preference but rather on the vain hope of "getting people out of their cars." 

Road construction receives a low priority in modern America, in favor of expanding mass transit and encouraging "active transportation" (biking and walking) by building more sidewalks and networks of bike lanes. In many of the largest metro areas during the last two decades, approved expansions of freeway capacity have been limited almost entirely to new carpool lanes. If people are going to stick with driving, the planners reason, we should at least push them to share rides by constructing preferential lanes for high-occupancy vehicles.

Congestion Pricing

When I first moved to Los Angeles from bucolic Santa Barbara in 1986, my easy five-minute commute was replaced by 30 to 45 minutes of L.A. freeway hell to go just 10 miles. Certain that there had to be a better way, I discovered the work of a small band of transportation researchers who were convinced that the missing ingredient in bringing freeway demand into sync with capacity was market pricing. Pioneered in the United States by Columbia University economist William Vickrey, who won a Nobel Prize in 1996, road pricing (later known as congestion pricing) had languished due to the lack of a workable method to charge variable prices. A New Mexico company called Amtech was in the process of addressing that gap by developing the first "toll tag" transponder. 

Even with this new technology, I could not imagine tradition-bound state transportation agencies using it to charge market prices on freeways. Yet for several decades, high-quality toll roads had already been financed, built, and operated by investor-owned companies in France, Italy, and Spain. The governments there would grant a company a long-term franchise, similar to those used for investor-owned electric utilities in the United States. Based on the franchise (called a "concession"), winning bidders could raise the capital to build the toll road, repaying their investors from toll revenues. So my proposal, in a 1988 policy paper published by the Reason Foundation, the 501(c)(3) nonprofit organization that publishes this magazine, was to invite the private sector to finance, build, and operate market-priced lanes on Southern California's congested freeways.

That paper soon came to the attention of Gov. George Deukmejian and California Department of Transportation (Caltrans) Director Bob Best, and in 1989 they secured passage of legislation to permit up to four pilot projects based on the idea. The first project to be developed was the 91 Express Lanes, four new variably priced lanes in the wide median on 10 miles of the highly congested SR 91 freeway in Orange County. The lanes, built for just $130 million in 1995 dollars, opened to traffic in December 1995, allowing commuters who had purchased transponders to decide whether higher speed was worth the congestion-sensitive price listed at the entrance of the passage.

 Although derided by skeptics, the tollway proved hugely popular. The demand for improved mobility was high enough that the Express Lanes have fully covered construction costs (via annual debt service payments on long-term toll revenue bonds) as well as all ongoing operation and maintenance expenses.

The 91 experiment demonstrated four important things. First, the long-term toll concession model from Europe is transferable to the United States. Second, many people are willing to pay for faster and more reliable rush-hour trips in highly congested freeway corridors. Third, transponder technology is a practical way of implementing congestion pricing. Fourth, variable, demand-based pricing is effective in keeping priced lanes flowing freely at the speed limit, even during the busiest rush-hour periods.

Challenging the Carpool-Lane Model

At first the success of the 91 Express Lanes did not stimulate much private-sector investment in congestion-relief lanes. But it did encourage the conversion of high-occupancy vehicle (HOV) lanes into something else the Reason Foundation invented: high-occupancy toll (HOT) lanes. By the early 1990s, research had found that HOV lanes make sense in only very limited conditions. Most such lanes—restricted to buses, vanpools, and carpools—were seriously underutilized. Commuters stuck in congestion in regular lanes fumed at the empty road to their left. And in a few cases in very congested metro areas, HOV lanes were jammed, to the point where they became nearly as backed up as regular lanes. The remedy for the latter problem was supposed to be increasing the required occupancy from two people to three—but that proved very difficult politically and therefore was rarely adopted. Thus, most HOV lanes were (and are) either too empty or too full.

Nearly all HOV lanes were built with federal money, making it illegal to convert them back to regular lanes despite their unpopularity. (The only known case of such a conversion, involving two HOV facilities in New Jersey, was authorized by a provision that a few of the state's representatives in Congress slipped into an unrelated 1998 bill.) So why not make lemonade out of these lemons, we proposed in 1993, by selling the excess capacity in underutilized HOV lanes to willing buyers? During the next 15 years, about a dozen HOV lane facilities were thus converted to HOT lanes in metro areas ranging from San Diego to Minneapolis to Miami, with each conversion adding to the evidence that commuters welcome having the choice between congested free lanes and uncongested pay lanes.

This evidence was not lost on those interested in investor-owned, toll-financed projects similar to the 91 Express Lanes. The second major such project, proposed to the Virginia Department of Transportation (VDOT) in 2002, was aimed at adding express toll lanes to the congestion-choked Capital Beltway in the Northern Virginia suburbs of Washington, D.C.

The Beltway is known to many locals as "the world's largest parking lot." Motorists desperately wanted relief, but the only plan VDOT had was a $3 billion project to add two old-style HOV lanes in each direction on the most congested 14-mile portion (the southwest quadrant) of the Beltway. There were two big problems with this plan: It faced enormous opposition because it required the condemnation of more than 300 homes and businesses to widen the Beltway, and VDOT didn't have anything close to $3 billion on hand.

Thinking Outside the Box

Fortunately, the Virginia legislature in 1995 had enacted the Public-Private Transportation Act, which authorized long-term toll concessions. Like California's 1989 law authorizing pilot toll-road projects, the Virginia law allowed private companies to make unsolicited proposals, rather than merely responding to VDOT requests. In 2002 one of America's leading engineering and construction firms, Fluor, submitted a proposal to fix the congestion problem on the Beltway. Rather than building tax-funded HOV lanes, Fluor proposed toll-funded HOT lanes. And rather than massively widening the Beltway, Fluor suggested an approach that would still add two lanes in each direction but would require only six takings of private property, mostly to accommodate rebuilt bridges and on/off ramps. The cost was estimated at $1 billion (one-third the cost of VDOT's plan), to be financed based on toll revenues. This was the 91 Express Lanes on steroids.

VDOT responded cautiously but positively to the proposal, beginning what ended up being several years of discussions about "design exceptions" that Fluor wanted in order to reduce the project's cost and footprint. Meanwhile, the company pitched the project to dozens of business and community groups in the Virginia suburbs. As Gary Groat, who was then Fluor's project development director, explained to me, federal rules (which had to be followed, since the Beltway is an interstate highway, I-495) forbid a state transportation department from "taking sides" on a proposed highway project while it is in the environmental evaluation phase. But no such restriction applies to the private-sector proponent. So Groat organized presentation after presentation, explaining HOT lanes and congestion pricing, the relatively low cost of the project, and the proposed financing based on toll revenue, all of which would make it possible to build the project in the near term (unlike VDOT's unaffordable HOV lanes). Most of all, Groat stressed the huge reduction in property takings entailed by Fluor's plan. These presentations, over several years, turned what had been strong opposition to the HOV lanes plan into wide support for the HOT lanes alternative.

As negotiations with VDOT progressed, the agency required many changes that would raise the project's cost; in addition, construction costs had increased significantly during the five years of review and negotiation. With the price tag now pushing $1.9 billion, it became apparent to Fluor that toll revenues alone would not pay for the project. This realization led to two key changes. First, the company teamed up with the Australian toll road company Transurban as both an investor and the planned operator. Second, the deal became a public-private partnership in which VDOT would contribute $409 million to cover most of its changes while Fluor/Transurban would finance the remaining $1.5 billion based on toll revenues. The deal was finalized in 2007, and the "financial close" took place in December of that year. It is noteworthy that investors were willing to commit to the project during the early months of the credit market crunch, which suggests the power of the financial model for toll lanes.

Construction began in spring 2008, and the project is scheduled to open to traffic by late 2012. Relief from rush-hour gridlock is finally in sight for hard-pressed Beltway commuters. Furthermore, the impending reality of the Beltway HOT lanes has stimulated serious research and planning for what could end up being a whole network of market-priced toll lanes in the D.C. metro area. Fluor/Transurban is negotiating a second project, which would convert 59 miles of the existing reversible two-lane HOV facility on I-95 approaching D.C. from the south into HOT lanes. 

Commuters in the D.C. metro area waste, on average, more than $1,500 a year in time and fuel due to congestion. The overall metro-area congestion cost in 2009 was $4 billion, eight times as high as the corresponding cost in 1984 when adjusted for inflation. So it's a promising sign that the Metropolitan Washington Council of Governments is refining plans for a region-wide network of congestion-priced lanes that could become an official part of the region's long-range transportation plan within the next year or so. 

Dallas and Fort Worth Join the Parade

Although policy makers in Texas have been more inclined to add highway capacity than their counterparts in many other parts of the country, the state has grown so much during the last 25 years that its highway fuel-tax funds have not kept pace. Consequently, the Dallas/Ft. Worth metro area ranks fifth in the nation in congestion cost, at $3.7 billion per year, with 66 percent of rush-hour travel occurring in congested conditions. Early last decade the Texas legislature passed a sweeping transportation public-private partnership act, aiming to tap into toll-based private capital to add needed capacity, especially on urban freeways.

The most congested part of Dallas has long been the LBJ Freeway (I-635). It constitutes the northern and eastern portions of an inner beltway (or "loop," in local lingo) around the city. When the LBJ was last expanded, the project was controversial due to the high-value properties along much of the corridor. In response, the Texas Department of Transportation (TxDOT) promised this would be the last widening of the freeway's footprint. But with continued growth in the metro area, the LBJ quickly ran out of capacity.

Because of its commitment not to expand the freeway's footprint, TxDOT approved a plan to add LBJ HOT lanes—up to three in each direction—by building tunnels beneath the existing freeway. When the agency issued a request for proposals, however, it took the open-minded approach of allowing bidders to propose alternatives. Of the two finalists, one went with the tunnels idea, at an estimated cost of $3.3 billion. But the winning bid, from Cintra/Meridiam, proposed an alternative that would cost only $2.6 billion. Via very creative engineering, the consortium proposed building most of the new toll lanes below the level of the main freeway lanes, with the main lanes cantilevered over the depressed express lanes. While in some ways more complicated to build, this approach produced significantly more roadway per dollar. 

As with the Beltway project in Northern Virginia, what TxDOT required—including full reconstruction of existing lanes and maintenance of the whole project for 52 years—proved more costly than the projected toll revenues could finance. The department therefore offered up to $700 million in state highway funds up front. But the winning bidder requested only $496 million, so TxDOT got the lanes it wanted while saving more than $200 million for other projects. And since Cintra/Meridiam had to finance only $2.6 billion, what the consortium needs to recover in toll revenues is significantly less than what tunnels would have cost.

This project also was financed during the credit market crunch, in June 2008. The tax-exempt private activity bonds received an investment-grade rating. Cintra, Meridiam, and the local fire and police pension fund together invested $665 million in private equity—an indication of investor confidence in the robustness of projected toll revenues over the 52-year concession period. The project is now under construction and expected to open to traffic in 2015.

Only six months earlier, the same Cintra/Meridiam team reached financial close on a slightly less ambitious toll-lanes project on the other side of the metro area: the $2.1 billion North Tarrant Express. This project will double the capacity of a 13-mile stretch of I-820 and SH 121/183 on the northeast side of Fort Worth, providing access from there to the huge DFW Airport. On I-820, the four new toll lanes can fit into the median, between the existing regular lanes. But on SH 183, some of the new lanes will be elevated, due to much less available right of way.

Are We There Yet?

We are on the verge of a paradigm shift in transportation. The ultimate manifestation of this re-think might be that investor-owned limited-access highways become a new category of utility, analogous to America's largely investor-owned electricity, natural gas, and telecommunications sectors, and the partially investor-owned water utility sector.

Steve Lockwood, a senior policy official at the Federal Highway Administration during the George H.W. Bush administration, helped build the case for loosening federal restrictions on tolling and pricing in the 1991 reauthorization of the federal highway and transit program. In a series of talks over several years, Lockwood made the analogy between highways and other network utilities, and even suggested that in the 21st century state transportation departments might evolve into public/private "transcorps" that would operate as businesses, more like state turnpike agencies than typical departments. The concept of investor-owned highways was independently fleshed out by former World Bank transport economist Gabriel Roth in his 1996 book Roads in a Market Economy.

Despite America's 19th-century history of privately franchised toll roads, most people have trouble seeing highways as businesses. If they have been exposed to economics, they think of roads as classic examples of "public goods," which must be provided by government because there's no way to exclude those who won't pay for them. While that might be mostly true for city streets and boulevards, it is not a problem for limited-access highways, where entry can only occur at certain points and charging for use is easy thanks to 21st-century electronic tolling, which is eliminating toll booths altogether. Just as electric utilities can charge variable rates to encourage customers to not use power-hungry appliances during periods of peak demand, expressway companies can charge more during rush hour to encourage motorists to shift nonessential trips to off-peak times.

In light of ever-worsening congestion, America's urban freeways constitute a massive case of government failure. They do not give their customers reliable mobility, and their method of financing (via flat-rate fuel taxes) vastly undercharges urban users (especially during rush hours) and generally overcharges rural highway users, whose roads cost far less to build. And without market pricing, we don't know how much urban expressway capacity people would be willing to pay for.

The private sector is capable of taking on the challenge of reinventing America's freeways. In recent years, several hundred global infrastructure investment funds have been created by investment banks, fund managers, and pension funds. As of 2011, Infrastructure Investor reported, the 30 largest such funds had raised $183 billion during the previous six years. The financial firm Probitas Partners estimated that all such funds had raised $200 billion by then. That kind of money is intended as equity investment, covering perhaps 25 percent of a project's cost, with the balance financed through bonds and other forms of debt. Thus, if these funds reach a total of $250 billion, that should make possible $1 trillion of infrastructure project investment, which could help expand and modernize a whole lot of congested, aging freeways.

Another factor supporting a paradigm shift is the increasing obsolescence of the fuel tax as America's primary highway funding source. The rising price of oil has motivated both business and government to seek alternatives to gasoline- and diesel-powered vehicles. Government-mandated increases in fuel economy have already reduced the yield of per gallon fuel taxes, since people today can drive twice as far on a gallon of gas as they did 25 years ago. By 2025 average fuel economy is mandated to double again, which will produce gaping shortfalls in highway-dedicated revenue (even if politicians were to stop siphoning off highway money to pay for bike paths and streetcars). 

We will have to shift to a more direct means of paying for highway use, and some form of per-mile electronic toll charge is almost certainly the best way forward. As the reality of this change sinks in, people may be able to get their minds around the concept of paying their monthly highway bill, just as they pay their monthly electricity, gas, and water bills.

Chronic freeway congestion ought to be intolerable to Americans. It's a crude form of rationing, based on time instead of money, like Soviet bread lines. The idea that everyone should be equally miserable, stuck in congestion, is similarly un-American. HOT lanes and toll concessions have given us a glimpse at a better future. 

Robert W. Poole Jr. is director of transportation policy at the Reason Foundation.