- Coping with reduced federal funding
- Managed arterials: using pricing for arterial congestion
- BRT gaining support
- More on availability-pay concessions
- Little change in commuting patterns
- Further thoughts on autonomous vehicles
- Upcoming Conferences
- News Notes
- Quotable Quotes
Regardless of what Congress does about the January 1st “fiscal cliff,” serious transportation analysts are coming to grips with the reality that “business as usual” will not be an option for federal transportation funding. The looming insolvency of the federal government will soon put an end to the practice of using general fund money (“borrowed from China”) to maintain spending levels beyond what comes into the Highway Trust Fund from highway user-tax revenues. Two recent reports from DC-based think tanks deal with several aspects of this funding crunch and make recommendations for reform of the federal program.
The first of these is “The Consequences of Reduced Federal Transportation Investment,” a joint product of the Eno Center for Transportation and the Bipartisan Policy Center. It crunches the numbers to estimate the impact on each state and on urban transit systems if the federal program were cut by 35%, the approximate amount of recent general-fund additions to the Trust Fund. In terms of highway capital investment, it identifies ten states with the highest federal share-not only small ones like Rhode Island (115%) and Vermont (75%) but also much bigger ones like Virginia (86%) and Minnesota (67%). For transit systems, the report identifies those most impacted by reductions in operating costs (whose federal share ranges from 15% to 21%) as well as the six most-dependent on federal capital grants (for which the federal share is currently between 81% and 100%). Faced with 35% cuts, the states and metro areas would very likely seek both to cut costs and increase state and local funding. The authors estimate that for highways, states would likely make up between 54% and 80% of the lost federal money. For transit, they are more pessimistic, assuming that no increases in state aid are forthcoming and that local governments would be unable or unwilling to cover more than a fraction of the lost federal aid. (http://bipartisanpolicy.org/sites/default/files/BPC-Eno%20Transportation%20Report.pdf)
The other report, “Highway Robbery,” is from the Center for American Progress. Its main focus is to call for an end to two current policies that adjust the redistribution of federal user-tax revenues: equity bonus and minimum set-aside rules. The authors provide a tabulation of how much each state gains or loses from these two rules. The equity bonus program was created to ensure that each state gets back at least 95% of the dollar amounts it collects and deposits in the Trust Fund (i.e., it reduces the extent of redistribution among states). CAP characterizes this as a bad thing, assuming that the states that benefit from equity bonus “paid a significant amount into the Highway Trust Fund but could not demonstrate sufficient need to get that money back through the need-based formulas” used to allocate most of the HTF monies. Equity bonus in 2010 accounted for $9.6 billion out of the $43.1 in the federal program. (www.americanprograss.org/wp-content/uploads/2012/09/CooperRoadsReport-5.pdf)
Both CAP and the Eno/Bipartisan authors dislike divvying up the federal spending by formula and would much prefer that the federal program be largely “performance-based.” Under this approach, the US DOT would set national goals for surface transportation, and all user-tax revenue would be allocated (in a mode-neutral way) to projects and programs judged by the feds to contribute most to achieving the federal goals. Most state DOTs, by contrast, view the federal program as an additional source of funding for what has always been a state-defined program, responding to state and local priorities. And highway user groups also tend to support the latter view, seeing “mode-neutral” funding as taxing highway users (cars and trucks) to pay for ever-expanding non-highway programs.
I’m with the highway users on this. But where does that leave us in terms of reforming the federal program to cope with up to 35% less revenue? My Reason colleague Adrian Moore and I argued in our 2010 policy study that it is time to refocus the federal program on what is truly federal: interstate commerce. That approach would devolve to the states responsibility for all non-highway programs and all highway programs except the Interstates and the National Highway System. (https://reason.org/studies/show/highway-trust-fund-reform)
To help states cope with their added responsibilities, Congress could do the following:
- Remove the remaining federal restrictions on tolling;
- Continue the TIFIA loan program at its newly expanded size and remove the current $15 billion cap on tax-exempt private activity bonds for highway and transit projects;
- Eliminate the latest increase in federal fuel-economy standards (54.5 mpg by 2025), which ARTBA estimates would avoid losing tens of billions in Trust Fund fuel tax revenues between now and 2025.
How could metro areas and transit agencies cope with this reconfiguring of the federal role? To begin with, they could make more aggressive use of managed lanes and managed arterials (see next article). In both cases, the reconfigured freeways and arterials would not only provide congestion relief for drivers; they would also provide uncongested guideways for region-wide express bus service. Second, transit agencies could make sensible decisions to make much greater use of buses and bus rapid transit (BRT), which is far more cost-effective than light rail and streetcars. Third, they could make roadways work better via more aggressive use of traffic signal coordination and freeway ramp metering. Fourth, they could more actively promote vanpools and telecommuting centers, both of which are highly cost-effective.
It’s incorrect to assess transportation’s future as if the only thing that will change is the amount of federal assistance. “Business as usual” must also change on the spending side, with a much greater emphasis on getting more bang for the bucks.
Over the past 15 years or so, the transportation world has increasingly embraced managed lanes-variably priced lanes on expressways that enable drivers to pay for faster and more reliable trips. As metro areas move toward integrated networks of expressway managed lanes, serious congestion relief will be achieved on that portion of the region’s roadway system. But in most metro areas, arterials carry a large fraction of peak-period trips, and many of those roadways are similarly congested during peaks. And as with freeways, it can be very costly and politically difficult to condemn land and widen major arterials. Wouldn’t it be nice if there were a way to apply the managed lanes principle to such arterials?
That was the challenge Chris Swenson and I faced in 2009 as he and I worked on a congestion-relief study for Lee County (Fort Myers), Florida. Without a freeway system (and with no support for creating one), it was obvious that any serious reduction in congestion would have to deal with the area’s major arterials. As a traffic engineer, Chris pointed out to me that when it comes to arterials, the biggest constraint on vehicle throughput is signalized intersections, especially in southern Florida where cycle times can exceed three minutes. Based on a 2003 project on which he had worked under a federal Value Pricing grant, Chris suggested adding grade separations at major signalized intersections. He was able to show that adding such features to a six-lane arterial would increase vehicle throughput significantly more than widening it to eight lanes. He also came up with the idea of charging a modest electronic toll for each use of the overpass or underpass, thereby helping to pay for this infrastructure improvement.
We fleshed out the concept further in the Reason Foundation’s 2012 policy study, “Increasing Mobility in Southeast Florida,” which proposed a network of such “managed arterials” in Miami-Dade, Broward, and Palm Beach Counties in addition to a three-county managed lanes network. (https://reason.org/studies/show/mobility-in-southeast-florida) We also described managed arterials in a paper presented at the 2012 Transportation Research Board Annual Meeting, which will soon appear in TRB’s Transportation Research Record. In the system we sketched out for Southeast Florida, our assumptions on pricing and usage led to an estimate that toll revenues from the managed arterial system could cover about 75% of its construction cost.
The idea of grade separations at major arterial intersections goes back a long way. In “Planning for Cars in Cities,” Jeffrey Brown, Eric Morris, and Brian Taylor discuss the 1924 document, “A Major Traffic Street Plan for Los Angeles” by F. L. Olmsted, H. Bartholomew, and C. Cheney, which called for grade separations at particularly busy intersections. (Journal of the American Planning Association, Spring 2009, Vol. 75, No. 2)
Is anyone taking this idea seriously? Lee County, FL, where the idea originated, had planned to convert its busiest arterial (Colonial Blvd.) into a managed arterial, prior to the housing market collapse, before which it had been one of the nation’s fastest-growing counties. Those plans are still on hold. In Miami-Dade County, a project development and environmental study is under way on converting the South Dade Busway into what amounts to a managed arterial, with grade separations at many intersections and the roadway open to paying customers in addition to buses. The Florida DOT announced in August that it is studying a possible project in Jacksonville to build a tolled flyover connecting I-95 with Butler Blvd. Without toll revenues, the $125 million congestion-relief project would not get built. And a “bus toll lanes” study is proceeding in Tampa, a joint project of the Tampa Hillsborough Expressway Authority and Hillsborough Area Regional Transit. Since I know of no projects akin to these in any other states, I’m betting on the first application of the managed arterial concept occurring in Florida.
A recent article in Governing magazine reported growing implementation of bus rapid transit (BRT) projects around the country. Some of this is occurring in mid-size metro areas such as Hartford, Jacksonville, and Nashville, while other projects are being implemented in cities with mature transit systems, including Chicago, New York, and San Francisco. New York already has three BRT lines and another three in the planning stages; it eventually aims to have 16 of them. (www.governing.com/topics/transportation-infrastructure/gov-bus-rapid-transit-gaining-traction-despite-concerns.html)
In July the GAO produced a very useful study on the subject: “Bus Rapid Transit: Projects Improve Transit Service and Can Contribute to Economic Development” (GAO-12-811). The GAO analysts studied 20 BRT projects completed since 2005. Most of these were what some dub “BRT-Lite,” in that they do not make use of exclusive lanes or fancy stations. But thanks to omitting those costly features, nearly all of these projects cost far less than either a “BRT-Heavy” project or a light-rail or streetcar alternative. And despite lacking those features, most of the 20 projects produced significant (from 20% to 80%) gains in ridership and significant travel-time savings (15% to 35%).
The Federal Transit Administration provides grant funding for BRT under both its New Starts and Small Starts programs. Although it’s widely believed that to qualify for the larger grants available from New Starts a BRT project must have a separate right-of-way for a “substantial” portion of its distance (at least during peak periods), that is not actually the case, GAO points out. A BRT project can still qualify if it “represents a substantial investment in a defined corridor as demonstrated by features such as park-and-ride lots, transit stations, bus arrival and departure signage, intelligent transportation systems technology, traffic signal priority, off-board fare collection, advanced bus technology, and other features that support long-term corridor investment.” Those are worthwhile features, and it’s good to see that FTA is not insisting on exclusive lanes, which in most cases are neither politically nor operationally feasible, due to the large negative impact on traffic congestion if an existing lane is converted to bus-only.
Is an exclusive bus lane a feature or a bug? Proponents argue that a community will only gain economic development benefits from BRT if it is seen as permanent, just like a rail line. But the opposing view sees the flexibility of BRT as a virtue, both tactically and strategically. In the short term, as GAO notes, if trouble occurs on the roadway (major accident, temporary street closure, etc.), a BRT vehicle can route around it, which a rail vehicle cannot do. Longer term, the fact is that cities evolve, and land-use patterns change. A transit system based on buses, including BRT, has the flexibility to adapt to such changes at low cost, compared with rail transit.
What still irks me about most BRT reports (including this one) is the short shrift they give to the synergy between priced lanes and BRT. For BRT-Heavy, which tends to be used for longer trips (including region-wide express service), a network of priced managed lanes on the freeway system provides the virtual equivalent of exclusive bus lanes. And what my colleagues and I have dubbed Managed Arterials (electronically tolled bypasses of signalized intersections) can do the same for BRT-Heavy on major arterials. The Reason Foundation’s 2012 Southeast Florida mobility study laid out an integrated plan for BRT-Lite on many arterials and BRT-Heavy on a network of managed lanes and managed arterials.
BRT is a good idea, but it can be even more effective if coupled with priced lanes and flyovers.
Last month’s article on “the perils of availability-pay concessions” generated a number of responses. Since some of these comments were made in confidence, I am withholding names and affiliations.
One concession company official agreed with my assessment, and offered some useful perspective from their U.S. experience. When they bid on an availability-pay project, he wrote, “the traffic numbers are not that relevant, since deviation of this traffic will [only] marginally affect operating expense and capital expenditure, and therefore will not really affect the internal rate of return. [For the] same reason, no one that is bidding has an incentive to optimize the scope of the project to the actual needs of the corridor.” Hence, “by taking away traffic or demand risk from the private sector, the reality check or economic viability sense-check, that is always needed for any public infrastructure, is gone.”
Another concession company official agreed that the both the need for more revenue for transportation and the incentive effects of transferring traffic and revenue risk are important advantages of revenue-risk concessions. They are well-suited to new (green-field) toll roads and bridges, as well as major additions such as managed lanes. But he suggested that for projects to replace worn-out bridges, there are two cases where availability-pay concessions might make sense. First is the case of a bridge that is essential but whose traffic levels will not support even a majority of the costs, if tolled. Second is cases where both the existing and new bridge are tolled but the public agency has strong policy reasons for setting the toll rates and using the revenue.
A third response came from a public agency that is going forward with a new toll bridge but is planning an availability-pay concession structure. His reasons for this choice are the following:
- The concession company’s cost of capital will be lower because it is not taking the revenue risk, which remains with the government agency. And that means the availability payments can be lower.
- Any toll revenue above what the agency needs for availability payments can be used for other Title 23-eligible transportation projects (or maintained as a liquidity reserve for the project).
- The concession structure still transfers construction and operations & maintenance risk to the company, and they will still be motivated to minimize life-cycle costs in their design-all important benefits of concessions.
- The company will be doing the toll collection, ensuring efficiency, and the toll rate setting is protected from political pressure by the bond trust agreement.
Those look to me like valid points, though I don’t think point 3 fully addresses the lack of incentives for “optimizing the scope of the project,” as noted by the first commentator. I still think that is one of the most important benefits of revenue-risk concession structures: they go much further towards making a highway or bridge into a true business venture.
The 2011 commuting data from the American Community Survey were released last month. They showed almost no change from 2010, but continued to be quite a bit different from 2000. That made the USA Today headline from Sept. 20th seem pretty bizarre: “Fewer Americans Commuting Solo.”
That’s not what the data actually show. Drive-alone commuting went from 76.57% in 2010 to 76.40% in 2011-essentially a rounding error. More relevant is the increase in drive-alone commuting since 2000, when the figure was 75.7%. Carpooling was essentially unchanged between 2010 and 2011, but has declined from 12.19% in 2000 to just 9.68% in 2011. Transit’s share increased slightly, from 4.57% in 2000 to 5.03% in 2011. The most significant increase was in telecommuting (work at home), which grew from 3.26% in 2000 to 4.34% in 2011. Thus, more people telecommute than the total that walk or bicycle to work (3.37%).
While it’s true that “the dismal economy and skyrocketing gas prices” are affecting some commuters at the margin, what amazes me is that solo commuting is holding up so well, given those economic factors. That’s probably because it’s hard to get to most jobs in a reasonable amount of time via transit. A Brookings Institution analysis that I reported on last year analyzed data from 100 metro areas. It found that only 7% of workplaces could be reached by transit in 45 minutes or less-yet the average commute by auto takes only 26 minutes. To reach even 13% of workplaces by transit requires 60 minutes, according to the Brookings analysis. For those willing to spend 90 minutes each way, 30% of jobs are within reach. And with transit budgets under serious pressures for cutbacks, don’t look for much transit expansion in coming years.
My somewhat critical assessment last month of the KPMG report on autonomous vehicles stimulated some interesting feedback. Several readers pointed out that the $70,000 cost of the rotating LIDAR sensor on the Google car is the price in very small quantities. If and when autonomous vehicles are put into mass production, the unit cost of such a device could decrease by at least an order of magnitude. One reader cited a recent news article in which a German supplier, Ibeo, says it has a contract with an undisclosed automaker to supply LIDAR systems for 2014 at a unit cost of about $250. That would clearly make a major difference in the cost of equipment needed to make a vehicle autonomous.
Another article cited by a newsletter reader quoted Google’s Anthony Levandowski as estimating that once autonomous vehicles go into mass production, “an ordinary car could be retrofitted for a couple of thousand dollars.” But from a liability standpoint, retrofit is a whole different ballgame than cars designed and produced from the start as autonomous vehicles. Count me as highly skeptical of retrofitting.
One consultant who has written technical articles on this subject took issue with the KPMG report’s argument that vehicle-to-vehicle and/or vehicle-to-roadside communications (now abbreviated as V2X) will be a key enabler of autonomous vehicles. He tells me he is in touch with several companies that are pursuing sensor-only approaches to autonomy. V2X, he says, is not needed and is not on the critical path to vehicle autonomy. When and if it is implemented, automakers will use it to enhance performance but will not depend on it. These vehicles will operate on roads “as is,” in mixed traffic, he maintains. And he writes, “Automakers affirm this; there is no debate on this topic within the car industry.”
Note: I don’t have space to list all the transportation conferences going on; below are those that I am (or a Reason colleague is) participating in.
Keystone to Pennsylvania’s Transportation Future, Oct. 16, Hilton Harrisburg, Harrisburg, PA (Shirley Ybarra speaking) Details at: www.ncppp.org/calendars/Harrisburg_1210/HarrisburgPAflyer.pdf
Transpo2012, Oct. 28-31, Hyatt Regency Coconut Point, Bonita Springs, FL (Bob Poole speaking). Details at: www.itstranspo.org.
AASHTO Annual Meeting, Nov. 15-19, Westin Convention Center, Pittsburgh, PA (Bob Poole speaking on Nov. 18). Details at: www.aashtoannualmeeting.com.
PPP Proposed for Tappan Zee Bridge. The aging and structurally unsound Tappan Zee Bridge across the Hudson River needs to be replaced, but the new toll bridge is estimated to cost at least $5 billion. New York State has applied for a federal TIFIA loan, but how can the project be kept affordable and delivered on time? State Sen. Greg Ball announced last month that he will introduce legislation next year to permit public-private partnerships for major transportation projects. He made the announcement during a one-day workshop organized by the National Council on Public-Private Partnerships.
CBO Questions Electric Cars. A new report from the Congressional Budget Office asks some pointed questions about the wisdom of federal policies that subsidize plug-in hybrids and all-electric vehicles. The current $7,500 federal tax credit does not offset the much higher cost of these vehicles to purchasers. The costs to the government for the reductions in gasoline use and vehicle emissions are very high. And these policies will have the perverse effect of enabling producers to make and sell more of their high-gas-consumption vehicles than would be the case (under federal CAFE standards) than if these subsidies did not exist. “Effects of Federal Tax Credits for the Purchase of Electric Vehicles” can be downloaded from http://www.cbo.gov/publication/43576.
Low Price-Elasticity for Chicago-Area Toll Payers. On January 1st of this year, the Illinois Tollway put through an 87.5% increase in its toll rates for cars, to pay for a $12.1 billion expansion effort over the next 15 years. Many people expected toll transactions to decline significantly, but the Chicago Tribune reported (Aug. 29) that for the first half of the near, transactions decreased only 2.6% compared with the same period in 2011, and revenue is up by 44%. (The 87.5% figure averages 100% increases for cash customers and much smaller increases for transponder customers.) Truck tolls were not affected by this increase, having been raised in 2005 with another increase coming after 2015. During the first six months of this year, truck traffic was up by 5%.
Will Ohio Lease Its Turnpike? The Bond Buyer (Oct. 9th) reported that Ohio Gov. John Kasich will likely release a plan to privatize the Ohio Turnpike within the next 30 days. NPR’s “Marketplace” carried the story on Oct. 15th. The governor has told reporters that he is reviewing several alternatives for the Turnpike, developed under a consulting contract by KPMG. Kasich implied that unlike the Indiana Toll Road lease, for which 75 years of lease payments were received in an up-front lump-sum payment, if the Ohio Turnpike is leased, the payments will occur over time and would likely be reserved by the state for infrastructure needs.
Washington State DOT Exploring Road User Charges. The legislature last year directed the state’s Transportation Commission to “determine the feasibility of transitioning from the gas tax to a road user assessment system of paying for transportation.” The detailed assessment is being carried out by Washington State DOT, with the assistance of Cambridge Systematics, BERK, and D’Artagnan Consultants. An overview of the project is online at: http://socialcapitalreview.org/wp-content/uploads/2012/09/wsdot-ruc-presentation.pdf.
California Enacts Autonomous Vehicle Testing Law. Following the lead of Nevada and Florida, the California legislature enacted AB 1298, allowing autonomous (self-driving) cars to be tested on public highways. It gives the state’s Department of Motor Vehicles until Jan. 1, 2015 to set standards for such testing. Such vehicles may not be driven on public roadways until the DMV approves the manufacturer’s application; it also requires the vehicle to have a qualified human driver and $5 million worth of insurance.
More Toll Facilities Going Cashless. Three more toll roads in the Austin, Texas metro area will switch to all-electronic tolling as of January 1, 2013, according to Tollroadsnews (www.tollroadsnews.com/node/6180). The facilities that are converting are part of the Texas Central Turnpike System operated by Texas DOT. Three other toll roads in the metro area are already operating on a cashless basis. Separately, the Ambassador Bridge between Detroit and Windsor (Ontrario) switched to all-electronic tolling on October 1st. The bridge is the first in the nation to deploy new “multiprotocol” readers from 3M that can read every type of toll tag used in North America.
Connecticut’s Tolling Intentions. In last month’s article about Virginia’s current plan to use only one tolling point (near the NC border) for its program to reconstruct and modernize I-95 with toll finance, I mentioned similar discussions that have taken place in Connecticut. In response, I heard from Thomas Maziarz, chief of policy and planning at Connecticut DOT. He objected to my statement that his state would likely enact border tolls if it received permission to use toll finance for I-95 reconstruction. He noted that although border tolls have been discussed in the state, their two ongoing Value Pricing studies are “based on the assumption of a much broader-based tolling approach/coverage than border tolling.”
Follow-up re Xpress West Train to Las Vegas. In last month’s issue I wrote about the Reason Foundation study critiquing the proposed high-speed rail line from Victorville, CA to Las Vegas. Author Wendell Cox has just done a follow-up analysis analyzing the extent of weekend traffic congestion experienced by drivers on two relevant portions of the journey: Los Angeles to Victorville (where the train would begin) and Victorville to Las Vegas. For two recent weekends, the data show that the vast majority of the congestion is on the LA to Victorville leg. This reinforces the original study’s skepticism that large number of LA visitors to Las Vegas would drive to Victorville to board the train, in the interest of avoiding congestion on the trip. (https://reason.org/news/show/xpresswest-train-most-congestion)
“In the long run, the combination of shrinking federal funding for transportation and continuation of federal restrictions on tolling is not sustainable. Unless another robust revenue source is identified, tolling across all lanes of Interstates will be critical to finance the major Interstate reconstruction and rehabilitation that looms. Federal law still prohibits such tolling. . . . In my view, there is no sound policy reason for federal law restrictions on what facilities may be tolled, at least in urbanized areas where the predominant traffic is local and regional. Policy decisions on tolling are driven by state and local needs and local citizen input. Local transportation bodies and the elected bodies that appoint them are accountable to the voters directly affected by tolling policy decisions. Where there is accountability, the decisions will reflect the balance of local interests and needs. Federal law and policy should accommodate, rather than inhibit, those decisions.”
-Fred Kessler, Nossaman, “Interstate Tolling Policy Must Change,” Public Works Financing, September 2012, p. 15
“The government’s ethanol content requirements are a classic case of good intentions run amok, with unintended consequences. Designed to reduce carbon emissions by providing subsidies for using a renewable resource, the ethanol requirement quickly became a boondoggle for large corporate corn farmers in the Midwest . . . . The subsidies have ended, but the issue is likely to get a lot hotter if Congress raises the minimum ethanol content requirement for gasoline to 15 percent at a time when drought has slashed corn crop estimates by a third and already driven up fresh food prices and the price of all products produced with corn syrup. A growing number of congressmen are changing their minds on extending the ethanol requirement, and more than 150 House members have written the EPA requesting a waiver of the ethanol requirement.”
-Peter T. Leach, “Unintended Consequences,” The Journal of Commerce, Sept. 10, 2012
“China’s high-speed rail ambitions put the Ministry of Railways so deeply in debt that by the end of last year it was forced to halt all construction and ask Beijing for a $126 billion bailout. Central authorities agreed to give it $31.5 billion to pay its state-owned suppliers and avoid an outright default, and had to issue a blanket guarantee on its bonds to help it raise more. While a handful of high-traffic lines, such as the Shanghai-Beijing route, have some prospect of breaking even, Prof. Zhao Jian of Beijing Jiaotong University compared the rest of the network to ‘a 160-story luxury hotel where only 11 stories are used and the occupancy rate of those floors is below 50%.'”
-Patrick Chovanec, Tsinghua University (Beijing), “China’s Solyndra Economy,” Wall Street Journal, Sept. 11, 2012
“Inadequate investment, Keynes argued, would leave the economy deprived of demand and workers bereft of employment. But mal-investment, argued Friedrich Hayek, an Austrian economist, would leave the economy poorly coordinated and workers stranded in the wrong jobs. In the past year, the spirits of Keynes and Hayek have done battle for the minds of China’s policymakers. This month Andrew Batson of GK Dragonomics, a research consultancy in Beijing, argued that Hayek seems to be winning. China’s leadership is now keen to avoid the ‘Hayekian risk’ of wasted investment, he wrote, even if that increases the ‘Keynesian risk’ of inadequate demand and weak growth. This Hayekian risk looms large, some argue, because China has already invested heavily in everything it normally stimulates when trouble looms. They point to disturbing examples of ghost cities, white elephants, and bridges to nowhere. One example is the Jiaozhou Bay Bridge, which connects the Huangdao district of Qingdao with the rest of the city, two locations already connected by a tunnel. The bridge is the longest overwater bridge in the world, but only because it does not follow a straight line.”
-FreeExchange, “Hayek on the Standing Committee,” The Economist, Sept. 15, 2012