I wrote the attached article in response to a cover story in The Weekly Standard that seriously misrepresented not only my own views but also three policy issues that I have been working on for more than 20 years: HOT lanes, express toll lanes, and transportation public-private partnerships (PPPs). A sympathetic editor at the Standard assured me that they would run some kind of a response from me, but it turned out that their policy is not to run “reply” articles in the print magazine, so the game plan was for it to be posted on their website. But then it turned out that my piece was too long, so that the editor was to condense it, for my approval. Weeks went by with no further word, and—given apparent internal divisions at the Weekly Standard over how to handle this—I decided that my least-bad course of action is to post my full response here at Reason.org. I apologize to those who’ve been waiting for my response, but finally here it is.
The cover story of the April 14th issue of The Weekly Standard was an article by Jonathan Last called “HOT and Bothered.” It was a long and fairly detailed critique of “HOT lanes,” that correctly cited my role in helping to popularize the concept. But Last’s critique seriously misunderstands what is going on as transportation agencies struggle to implement new approaches to reduce America’s seemingly intractable highway congestion. In this article I focus mostly on three serious misunderstandings—about priced express lanes, about the role of the private sector, and about who supports and opposes these new approaches.
Misconception No. 1: What’s HOT and What’s Not
Last’s first major misunderstanding is to conflate two quite different types of priced lanes. The term “HOT” (High-Occupancy Toll) refers to freeway lanes that began as high-occupancy vehicle (HOV) lanes and were later converted to priced lanes that allow carpools to use them at no charge.
Northern Virginia’s Capital Beltway (I-495) is using Express Toll Lanes (ETLs), a considerably different business model. The ETL business model is to offer faster and more reliable trips, for a price, to anyone who finds that the value of getting somewhere reliably on time is greater than the added cost of the toll. Some ETLs charge all vehicles, but it is often politically expedient to provide free or discounted access to some forms of high-occupancy vehicles (buses, van-pools, and sometimes three-person carpools).
By contrast, the HOT lanes business model is basically to encourage ride-sharing just as HOV lanes do but to sell any excess capacity to paying customers. Most HOT lanesgive away most of their capacity to two-person carpools, in keeping with their HOV-focused business model. This applies to many HOV-to-HOT lane conversions, including those in Minneapolis, San Diego, and Seattle.
Last makes only passing reference to the Express Toll Lanes on SR 91 in Orange County, California, but I had a major role in proposing this then-revolutionary idea to the California DOT (Caltrans) back in 1988, as a potential investor-funded project. Enabling legislation was passed the next year, and a competition yielded four winning proposals, of which the 91 Express Lanes was one. Entirely privately financed, the new lanes opened in December 1995 and have delivered reliable 65 mph rush-hour trips for Orange County commuters for 18 years and counting, thanks to variable, demand-based pricing. Those Express Toll Lanes were actually the inspiration for many HOV-to-HOT conversions, even though many of those doing the conversions failed to adopt the commercial business model, instead allowing limited numbers of paying vehicles to “buy-in” to the former HOV lanes.
But in recent years that mindset has been changing. Two recent conversions of HOV lanes—in Atlanta and Miami—have adopted the Express Toll Lanes business model. In both cases, each state’s Department of Transportation increased the occupancy required for free passage from two to three, and required all carpools to pre-register. Those moves ensured that the large majority of vehicles are paying the tolls and are therefore affected by the variable pricing, which is what ensures that the lanes remain free-flowing during peak periods. A critical point is that this also makes Express Toll Lanes much better corridors for fast and reliable express bus service. Miami has seen express bus ridership triple since the converted 95 Express Lanes opened four years ago.
The new Express Lanes on the DC area’s Capital Beltway are the most recent example of Express Toll Lanes. Contrary to Last’s implications, these all-new lanes, added to the Beltway at a total cost of $1.9 billion, don’t take anything away from anyone. What they do is add an option to “the world’s longest parking lot” that was not there before—faster and more reliable trips, when people of all income levels want and need this option. Allowing carpools of three or more to use the lanes for free was one of the conditions developer Transurban-Fluor had to accept, to make the policy for special lanes the same on the Beltway as on I-95. Comparable $2 billion scale ETL projects are under construction in Dallas and Fort Worth, and will soon start construction on I-4 in Orlando, all done via private investment under long-term public-private partnerships.
Misconception No. 2: These PPPs Really Are Partnerships
Last’s second major problem is a failure to understand how long-term public-private partnerships (PPPs) for highway projects work. Someone reading only Last’s article would see this type of arrangement as a form of crony capitalism, yielding “privatized profits and socialized losses.” That is the opposite of reality. The private investors in this and similar long-term public-private partnerships are entirely at risk—for construction cost overruns, for late completion, and most of all for the consequences of over-optimistic traffic and revenue projections. There is no guarantee they will ever make a profit, and an ever-present risk that the project will lose money or even go bankrupt.
That kind of risk transfer is one of the major benefits of using this kind of PPP for mega-projects (which are notorious for cost overruns and over-optimistic projections of ridership). And this risk transfer feature is not just theory. In the nearly 20-year history of long-term highway PPPs in America, two such toll road projects have gone bankrupt (Camino Columbia in Laredo, Texas and the South Bay Expressway in San Diego). There were no taxpayer bailouts in either case, and the investors lost heavily. Motorists did not. The assets were purchased out of bankruptcy (at a fraction of their cost to construct) by public agencies, and the toll roads continued to operate in serving their customers. Toll rates were reduced, thanks to the asset being acquired at a lower cost. Two similar bankruptcies—of new investor-financed toll tunnels—have occurred in Sydney, Australia within the past decade, with the same result. Investors took the hit, not taxpayers, and the roads have remained in service under new owners.
Last also misrepresents how the Beltway Express Lanes came to be. For many years the Beltway was plagued by chronic congestion, which the Virginia Department of Transportation promised to fix. Their solution was a $3 billion project to add two HOV lanes each way to the Beltway. There were just two problems with this plan: it would have required “taking” over 300 commercial and residential properties, and VDOT had no idea where to obtain $3 billion. Thanks to Virginia’s Public Private Transportation Act, in 2002 a major engineering/construction firm—Fluor—gave VDOT an unsolicited proposal for a better solution. Instead of HOV lanes, it would build ETLs, and if VDOT agreed to a whole list of design changes, it could cut the project cost to around $1 billion and avoid all but a handful of property takes. And at least equally important, projected toll revenues would cover the billion-dollar cost.
While a radical departure from the VDOT status quo, the offer was too promising to ignore. Over the next four years, VDOT and the company engaged in lengthy discussions, especially over the design changes, while VDOT conducted the environmental review. In the end, because the agency wanted more extensive connector ramps, more bridge replacements, and other features (and due to construction cost inflation), the price tag increased to $1.9 billion—but the huge reduction in property takes was preserved. The latter point won over a majority of the local population, who had opposed VDOT’s $3 billion HOV plan. But because VDOT insisted on 3-person carpools getting free passage (as on I-95), the toll revenue would no longer cover the full project cost, especially at $1.9 billion. So to make the project toll-financeable, VDOT did two things. First, it agreed to cover $400 million of the increased project cost it had caused, reducing the amount to be financed to $1.5 billion. Secondly, it agreed that if the volume of three-person carpools ever exceeded 24% (highly unlikely), VDOT would pay the tolls for the extra carpools.
That was neither a subsidy nor a guarantee of profit. It is simply an illustration that with both parties willing to work together to reach an acceptable solution, the project could get built and operated by the private sector, taking on all the risks noted previously, including the possibility that traffic and revenue will never reach the levels assumed in their financial model.
Last also misunderstands the financing of projects like this. He implies that Fluor and its partner, Transurban, have only put in $350 million rather than the $1.5 billion I noted above. The $350 million is their equity investment (like the down payment on a home). The tax-exempt toll revenue bonds and the tax-exempt subordinated loan from the Federal Highway Administration are the debt portion, like the mortgage on a home. What Last fails to point out is that Transurban-Fluor is entirely at risk to pay off those loans from the toll revenues generated by the project. A requirement of the federal loan is that the primary toll revenue bonds be investment-grade. That means the project had to pass financial thresholds, not just appeal to politicians. In the event of a future bankruptcy, both the revenue bonds and the federal loan would have first priority for repayment; the equity providers would be last in line.
Also, Virginia did not “give away the land.” VDOT still owns the land, and the toll road company is essentially leasing it for the term of the long-term PPP. The huge, detailed contract between the parties requires that the new lanes be maintained in top shape and be returned to VDOT at the end of the term in excellent condition.
These points all serve to illustrate that PPP deals like this are truly partnerships. They are painstakingly negotiated and structured to be win-win-win—for Virginia and its taxpayers, for the private companies, and for the driving public.
One other point on this subject: The Transurban-Fluor team made a subsequent unsolicited proposal to VDOT to convert the HOV lanes on I-95 into ETLs that would smoothly interface with the ETLs it was building on the Beltway. Under the partnership legislation, any unsolicited proposal must be publicized and the same opportunity offered to other bidders. In the Beltway case, there were no other bidders, so VDOT began negotiating with the team that originated the idea. With the I-95 proposal, there was one other bidder, and VDOT selected the better proposal. No sweetheart deals here; just various companies deciding which projects to pursue and VDOT making a merit-based selection.
Misconception No. 3: Smart-Growthers, Suburbs, and Toll Lanes
I had to laugh when I got to Last’s portrayal of Express Toll Lanes as part of the agenda of Smart Growth groups who hate the suburbs and suburbanites. Anyone who follows highway policy knows that these groups oppose anything that adds capacity to highways and makes driving better for commuters. And many of these groups even oppose converting existing HOV or other lanes to HOT lanes. For example, in response to a recent proposal from the Washington, DC, city government for HOT lanes on the 14th Street bridge and on DC’s portions of I-295 and I-395, the Coalition for Smarter Growth’s Stewart Schwartz said, “The key thing to think about is that even if you are able to improve rush-hour movements because of HOT lanes, the single occupant or even the carpool vehicles will still have to drive on DC streets.” Opposing car travel is their point, not making freeways work better.
In fact, the region-wide network of ETLs that the DC region’s Transportation Planning Board has been studying would facilitate commuting from the suburbs to downtowns, and from suburb to suburb, by making use of the network configuration of the existing freeway system. That would make the DC metro area less congested, and would likely make car commuting from the suburbs less stressful and congested, because almost all of the ETL network would be new lanes. But unlike new general-purpose lanes which would—sooner or later—fill up and become congested, the new ETLs would remain uncongested over time, just as those on SR 91 in Orange County have for the past 18 years. And all the traffic that shifted onto the ETLs on any given day would no longer be in the free lanes, reducing congestion there below what it would otherwise be.
Other Concerns About Toll Lanes
The article also raises other concerns, some that would apply to any such projects and a few that are unique to this one.
Egalitarians tend to take the highest peak-hour toll rate and multiply that by the number of work days in a year, to estimate the cost of using an ETL every day to and from work. After that, they generally refer to the ETLs as Lexus Lanes (a term I appreciate Last not using), implying that only the well-heeled will use them. That assessment does a disservice to the ability of middle- and lower-income people to figure out what is in their own best interest.
A single working mother with two kids in day care could well face late fees in excess of a dollar per minute for arriving late to pick them up at the end of her work day. You don’t have to be a genius to figure out that paying a $5 or even $10 toll to avoid a $20 or $25 late fee is a rational choice. Yet with most congested freeways, no such choice is available. A plumber or electrician could easily find it worthwhile to pay an ETL toll to save enough time to get in one additional job per day. Likewise, a family on its way to the airport for a long-planned vacation would gladly pay a toll to bypass freeway congestion and get there in time to board their plane. Moreover, people who are paid hourly, rather than being on salary, generally have much less flexibility about their work hours, so they may find themselves needing to be sure they get to work on time, so as not to have their pay docked or get fired. Again, on most freeways these people have no choice of paying a toll to avoid bad consequences—but they do now on the Beltway.
These examples illustrate what we observe in practice on HOT and ETL facilities across the country. For example, the 91 Express Lanes have 170,000 FasTrak transponders in service, but on any given day only about 34,000 people use those lanes. Most people do not use these priced lanes every day, all year long. Most use them a few times a week, when they judge the timesavings and predictable arrival time of a particular trip worth the cost of the toll. And studies by several state DOTs find that the largest numbers of vehicle makes using priced lanes are ordinary cars. On the Beltway ETLs, the four highest-share vehicle makes are Toyota (17%), Honda (15%), Ford (8%), and Nissan (5%).
On a point specific to the Beltway lanes, Last complains that the signage for the Express Lanes “withholds” the information about how much time a motorist will save by opting to use the pay lanes. The signage problem is more complex than he realizes. The early ETL and HOT projects were very simple: the original 91 Express Lanes had an entrance on one end and an exit on the other; likewise, the initial seven miles of the I-95 Express Lanes in Miami. The Beltway Express Lanes are the most complex ones opened to date. If you get on at the I-95/395 interchange, there are nearly a dozen possible “trips” you could take between entry and exit. There is no way that electronic signs that must be read quickly while driving can convey all those possible outcomes. The best that can be done on the roadway is to list one or two destinations and the current price and time via Express Lane versus free lane. That is what is planned for the forthcoming I-4 Express Lanes in Orlando.
Finally, Last also contends that the jury is still out on whether adding a free (general purpose) lane to a congested freeway is better or worse than adding a priced lane. There’s been extensive research on this issue, as well as a growing amount of empirical data. In the short term, on a congested freeway, the rush-hour flow rate in the new lane will probably be higher than if the new lane is priced. But the real question is how sustainable that performance is. New capacity will attract back to the freeway drivers who got so fed up with its congestion that they opted for parallel surface streets instead. And that “induced demand” will tend to fill up the new lane rapidly, to the point of congestion.
All traffic engineers know that if you can keep the number of vehicles per hour in a lane to about 1,800-1,900, it will flow smoothly (though feeling a bit crowded) at pretty high speed. With much more than that number attempting to get into the lane, speed decreases as the flow rate increases to a maximum of about 2,000 vehicles per lane per hour. With any more cars than that, both flow and speed decrease into the awful stop-and-go conditions we all know and hate. During rush hours when demand for the freeway is very high, the actual flow rate can deteriorate to as low as 900 to 1,200/lane/hour, at speeds in the 10-15 mph range. But with variable pricing, the flow in the toll lanes can be kept at 1,800-1,900/lane/hour at 50-65 mph. There is no mechanism except variable pricing that can produce this combination of high flow and high speeds during rush hours—except a serious recession.
Nothing that I’ve presented here is new information. The story of how the Beltway ETLs came to be was recounted in my March 2012 Reason article, “Fixing the Freeways.” A number of Reason Foundation policy studies, more recent than the 1993 and 1999 ones Last cites, explain the traffic engineers’ speed vs. throughput curve and the rationale for long-term public-private partnerships.
I find it ironic and dismaying to see a free-market conservative defending the status quo model for U.S. highways—one with government ownership and operation, no pricing, funded by taxes rather than direct user charges, and with investment decisions made largely for political rather than economic reasons. My friend, former World Bank transportation economist Gabriel Roth, has called this model America’s socialized highway system. It’s a strange model for free-market conservatives to champion.
Robert Poole is Director of Transportation Policy and the Searle Freedom Trust Transportation Fellow at the Reason Foundation. He is working on a book on 21st century highway policy.