Traffic (52 page)

tendency of birds to breed: Harvard University’s Richard Forman, the dean of the “road ecology” movement, noted in a typical study that bobolinks and other grassland birds in Massachusetts do not breed when their nest sites are close to high-traffic streets (on streets with three thousand or fewer vehicles a day, they do breed). The suggested culprit is noise. See R. T. T. Forman, B. Reineking, and A. M. Hersperger, “Road Traffic and Nearby Grassland Bird Patterns in a Suburbanizing Landscape,”
Environmental Management,
vol. 29 (2002), pp. 782–800, and R. T. T. Forman, et al.,
Road Ecology: Science and Solutions
(Washington, D.C.: Island Press, 2003). See also J. A. Jaeger, L. Fahrig, and W. Haber, “Reducing Habitat Fragmentation by Roads: A Comparison of Measures and Scales, in
Proceedings of the 2005 International Conference on Ecology and Transportation,
eds. C. L. Irwin, P. Garrett, and K. P. McDermott (Raleigh: Center for Transportation and the Environment, North Carolina State University, 2006), pp. 13–17.

less able to afford cars: See Donald Appleyard, M. Sue Gerson, and Mark Lintell,
Livable Urban Streets: Managing Auto Traffic in Neighborhoods,
a report prepared for the Federal Highway Administration, 1976. Many of Appleyard’s findings were reconfirmed in a study by the New York City group Transportation Alternatives, “Traffic’s Human Toll,” 2006; available at
http://www.transalt.org/press/releases/061004trafficshumantoll.htm
.

were taxing the poor: It follows that poorer areas also suffer more exposure to the exhaust of passing traffic. Studies in Leeds, England, for example, found that economically disadvantaged areas had higher levels of nitrogen dioxide. See G. Parkhurst, G. Dudley, G. Lyons, E. Avineri, K. Chatterjee, and D. Holley, “Understanding the Distributional Impacts of Road Pricing,” Department of Transport, United Kingdom, 2006.

by Garrett Hardin: See Garrett Hardin, The Tragedy of the Commons.”
Science,
December 13, 1968.

oft-invoked “tragedy”: Shi-Ling Hsu, “What
Is
a Tragedy of the Commons? Overfishing and the Campaign Spending Problem,” February 21, 2005, bepress Legal Series, Working Paper 463;
http://law.bepress.com/expresso/eps/463
.

any traffic engineer: Gary Toth, a planner with the New Jersey Department of Transportation, told me in a conversation in early 2007: “We ran a calculation this week for the twenty congestion-related projects that I have in my division. Those twenty represent about ten percent of the congestion in New Jersey. The construction cost to fix those is $6.7 billion.” Given that about $100 million of the department’s $600 to $700 million budget can be spent on congestion projects, he said that “at the rate the public is providing funding for us,” he could expect those congestion projects to be completed in 670 years.

build new ones: See, for example, Joel Kotkin, “Road Work,”
Wall Street Journal,
August 28, 2007.

all those fuel taxes: Mark Delucchi of the Institute of Transportation Studies at UC-Davis estimates that current payments in the form of fees and taxes by car users to the federal government fall below the costs the federal government pays for car use by some 20 to 70 cents per gallon of fuel. See Mark A. Delucchi, “Do Motor-Vehicle Users in the US Pay Their Way?” Institute of Transportation Studies, Research Report UCD-ITS-RP-07-17, University of California, Davis, 2007.

in the 1960s: See “The Gasoline Tax: Should It Rise?”
Wall Street Journal,
August 18–19, 2007.

“90 percent of the time”: Martin Wachs, “Fighting Traffic Congestion with Information Technology,”
Issues in Science and Technology,
vol. 19 (2002), pp. 43–50.

two Canadian researchers: See K. Mucsi and A. M. Khan, “Effectiveness of Additional Lanes at Signalized Intersections,”
Institute of Transportation Engineers Journal,
January 2003, pp. 26–30. The authors also note that additions to larger intersections will become congested more quickly than additions to smaller crossroads. They write: “If a one-lane road (per direction) gets saturated at 1,000 vehicles per hour (vph) and annual growth is 3 percent, the additional lane will have an uncongested lifetime of approximately 24 years. If a three-lane road (per direction) gets saturated at 3,000 vph and annual growth again is 3 percent, the uncongested lifetime of the additional lane is only 10 years, even without factoring in the diminishing marginal capacity benefit of the additional lane. The diminishing capacity benefits of additional lanes only speed up the process.”

the fourth just 385: Engineers, for their part, have responded to the problems of large intersections by building highway-style overpasses, which are not just expensive but can look rather freakish rising out of an otherwise flat suburban environment, or with the so-called continuous-flow intersection, a breathtakingly complex creature that removes the left-turn conflict from the main intersection by having drivers turn left
before
they get to the actual intersection; this is a bit unnerving for some drivers, as the design makes it seem as if they are headed into the oncoming lane. Early studies, however, have shown that these designs actually move more traffic more safely than conventional intersections. At an intersection in Baton Rouge, Louisiana, wait times were reduced from four minutes to one. For a good roundup of CFI intersections, with animations, visit AMBD Engineering’s Web site at
http://www.abmb.com/cfi.htm
.

an estimated 12.7 percent: This number is taken from H. Teng and J. P. Masinick, “An Analysis of the Impact of Rubbernecking on Urban Freeway Traffic,” Center for Transportation Studies, University of Virginia, Report No. UVACTS-15-0-62, 2004, p. 47.

“it is a bad bargain”: Thomas Schelling,
Micromotives and Macrobehavior
(New York: W.W. Norton, 2006), p. 125.

photos of incidents: Melissa Leong, “Best and Worst: Driving GTA’s Highways with Sgt. Cam Woolley,”
National Post,
July 18, 2007.

“or other vehicles”: Andrea Glaze and James Ellis, “Pilot Study of Distracted Drivers,” Center for Public Policy, Virginia Commonwealth University, January 2003.

would have gone up: As a thought experiment, consider that the salad bar was actually free. What would happen? There would be huge queues of people lined up for the free food. As Tim Harford points out, “We recognize that food, clothes, and houses cannot be free or we would have quickly run out of them. It is because roads are free that we have run out of spare road space.” From Harford,
The Undercover Economist
(Oxford: Oxford University Press, 2004), p. 88.

more people want to use them?: William Vickrey, “Pricing in Urban and Suburban Transport,”
American Economic Review,
vol. 53 (1963). Reprinted in Richard Arnott, Kenneth Arrow, Anthony B. Atkinson, and Jacques H. Drèze., eds.,
Public Economics: William Vickrey
(Cambridge: Cambridge University Press, 1994).

the results to friends: The Vickrey story is taken from a working paper by Ron Harstad at the University of Missouri, available at
www.economics.missouri.edu/working-papers/2005/wp0519_harstad.pdf
.

rationalize its loss: For an interesting discussion of these ideas based on laboratory experiments, see Erica Mina Okada and Stephen J. Hoch, “Spending Time Versus Spending Money,”
Journal of Consumer Research,
vol. 31 (2004), pp. 313–23.

than on another day: Richard Clegg, “An Empirical Study of Day-to-Day Variability in Driver Travel Behavior,” Department of Mathematics, University of York, Heslington. Retrieved at
www.richardclegg.org/pubs/rgc_utsg2005.doc
.

dropped by 13 percent: Kitchen, in an e-mail, pointed out that all results are “non-equilibrium.” That is, if the roads were actually tolled, traffic speeds would improve, attracting additional users.

increase speeds by 50 percent: John D. McKinnon, “Bush Plays Traffic Cop in Budget Request,”
Wall Street Journal,
February 5, 2007.

jump by 5 percent: Philip Bagwell,
The Transport Revolution
(London: Rout-ledge, 1988), p. 375.

go into buses: As Puget Sound’s Kitchen points out, the revenues generated from economically efficient tolling are greater than the total surplus that is gained through drivers’ saved time, which makes the question of how revenues from pricing get redistributed an important, if often neglected, one.

thus more popular: For more on this “virtuous circle,” see Kenneth A. Small, “Unnoticed Lessons from London: Road Pricing and Public Transit,”
Access,
vol. 26 (2005), pp. 10–15.

show up so often in networks: An interesting example from the traffic world that recalls Laval’s monorail case is Route 29 in Trenton, New Jersey. A product of the 1960s era in which cities elected to build massive high-speed highways through the middles of downtowns or alongside waterways, Route 29 is a dangerous road, with numerous crashes and some two fatalities over a fifteen-year period, as I was told by Gary Toth, an engineer with the New Jersey Department of Transportation. Part of the reason was that cars were “blitzing” down a road that was marked for 45 miles per hour but designed more like a 65-mile-per-hour freeway (with all the standard “safety” provisions of clear zones and the like). Drivers would then inevitably encounter the back of a queue of cars waiting at a signalized intersection; it was a classic “hurry up and wait” situation. Rather than have a bunch of high-speed cars encounter a single light with a long delay, Toth and his colleagues wondered what would happen if Route 29 was converted from a highway into a more aesthetically appropriate and pleasant “urban boulevard,” with a lower speed limit and several more sets of signalized crossings. Wouldn’t that just
cause
more congestion? Wouldn’t it foist an unconscionable delay upon drivers? When they ran simulations, they found that the new system added only two minutes to the total trip during peak times. Instead of one large queue at a signal, the wait would be redistributed among a set of lights. Importantly, the new system carries the added benefit of being much safer as well, as it involves less sudden braking at high speed.

“for Easter Sunday”: There are other strange dynamics at work; after running simulations, Laval rejected a plan to double the capacity of the Country Bear Jamboree. “People had the perception it was popular because it had such long lines,” he said. “It was really just because it had limited capacity. It’s a common misperception.”

because
it is expensive: See, for example, Daniel Machalaba, “Paying for VIP Treatment in a Traffic Jam,”
Wall Street Journal,
June 21, 2007.

as the toll goes up: As Moshe Ben-Akiva, director of the Intelligent Transportation Systems program at the Massachusetts Institute of Technology, described it to me, the challenge with dynamic pricing is that the price changes depending on your objective: “You may want to charge people for time they actually save. That will mean if congestion builds up on the toll road, you reduce the price. On the other hand, you may want to maintain a certain level of speed on the toll road. If congestion builds up you may want to
increase
the toll so as to not have stop-and-go traffic on the toll road. There is some confusion going on right now as to what strategy is best.”

by changing their plans: Ronald Koo and Younbin Yim, “Commuter Response to Traffic Information on an Incident,” September 1, 1998, California Partners for Advanced Transit and Highways (PATH), Working Papers: Paper UCB-ITS PWP-98-26;
http://repositories.cdlib.org/its/path/papers/UCB-ITS-PWP-98-26
.

have some information: In one experimental study, for example, eighteen subjects had to choose between two roads, one of which was faster only if an equal number of people chose the opposite road. The subjects would receive a higher payoff for successfully choosing the quickest route. As it happens, over the long run most people split evenly onto the two roads. But there were “daily” fluctuations, and, more important, these were still happening after two hundred trials. The reason is that people continually tried to outguess each other with better strategies (it turned out drivers did better when they simply chose the same route each time) or find out if the other road was in fact better. Interestingly, in a second trial, drivers were given information about the travel time of the route they did not take, meaning they did not have to change roads to know what the conditions were. The fact that drivers had this information had only a “small effect” on the fluctuation between the two roads from day to day. See Reinhard Selten, Michael Shreckenberg, Thomas Pitz, Thorsten Chmur, and Sebastian Kube, “Experiments and Simulations on Day-to-Day Route Choice-Behaviour,” April 2003, CESifo Working Paper Series No. 900; available at
http://ssrn.com/abstract=393841
.

flocked to the highway: Virginia Groark, “Dan Ryan Traffic Flow Changes by Minute—Like Chicago Weather,”
Chicago Tribune,
April 5, 2006.

shown the best routes: See Moshe Ben-Akiva, Andre De Palma, and Isam Kays, “Dynamic Network Models and Driver Information Systems,”
Transportation Research A,
vol. 25A, no. 5 (1991), pp. 251–66.

in two-way traffic: Sarah Murray, “The Green Way to Keep on Trucking,”
Financial Times,
March 13, 2007.

no longer rise tomorrow: Tim Harford,
The Undercover Economist
(Oxford: Oxford University Press, 2005), p. 138.

to the same problem: There is an interesting analogy in all this between traffic and the stock market. In theory, as individual investors are able to more closely track the real-time fluctuation of stock prices via the Internet, having access to more and more bits of information about companies, they should be better able to make informed decisions that more quickly translate into stock prices, and market volatility should go down (see Daniel Gross, “Where Have All the Stock Bubbles Gone?”
Slate,
January 3, 2006). But Brad Barber and Terrance Odean have suggested several potential problems that may arise as a result of many more people having access to inexpensive, almost-instantaneous stock trading via the Internet, including the availability of “faster feedback” that may prompt investors to focus too much on recent performance. Those trying to profit from short-term “momentum cycles,” they write, may actually increase volatility. (They note that individual stocks have increased in volatility over the past several decades, for reasons that they say are not well understood.) Brad M. Barber and Terrance Odean, “The Internet and the Investor,”
The Journal of Economic Perspectives,
vol. 15, no. 7 (Winter 2001), pp. 41–54.