How to break up traffic jams

As reprinted in Moore, Robert L. and James D. Whitney. Microeconomic Principles in Action. Englewood Cliffs, NJ: Prentice Hall, 1990, pp. 112-114. This article on highway congestion illustrates how common externalities actually are. The article deals with the issues of peak versus off-peak highway use and details some recent innovations in the metering of traffic. The questions include a quantitative problem on the external costs of congestion and cover additional issues, such as the nature of congestion, other places where it occurs, and metering innovations discussed in the article. How to Break Up Traffic Jams

By JONATHAN MARSHALL

From San Francisco to Houston to New York City, ordinary citizens worry less about the threat of nuclear war or the Latin debt crisis than whether they will emerge each day of sound mind and body from nightmarish traffic jams.

In the luckiest cities, billions of dollars invested in new highways, heavy rail systems and subsidized buses have bought only a few years' relief from inevitable gridlock. Elsewhere, policies of simply increasing the supply of transportation have proved to be wildly expensive failures.

Urban Americans can't afford to ignore the demand side of the equation any longer. No strategy to solve traffic woes will work unless individual drivers pay the full costs they impose on society. The answer isn't a punitive crackdown on private autos, but rather a fair and equitable levy on their use of the roads.

Politicians and transit planners have ducked the issue of road pricing as unrealistic and unworkable. But engineering solutions today make the economists' case for road pricing unanswerable. And government officials may change their tune. Last week, New York City's transportation commissioner, Ross Sandier, said that "dramatic" steps were needed to curb city traffic congestion. Among the possible proposals he mentioned: charging motorists fees for entering certain areas.

Randall Pozdena, a transportation specialist at the Federal Reserve Bank of San Francisco, observes that highway-user charges are "perhaps the most important issue in transportation policy. Transportation impacts land use, commerce, energy and the environment in very direct and important ways. To make sensible policy in these other areas, we must know what the highway system costs us and what the effects of changing our policy might be. Brute force policies in these other areas may be much less successful and more costly than a rational highway pricing mechanism."

The Law of Peak-Hour Congestion

Giving cars a free ride on highways is like giving desert dwellers free water for lawns, pools and golf courses: They'll use everything you give them. The Law of Peak-Hour Congestion follows: Traffic always grows to fill the available capacity.

Rush-hour auto commuters pay for only a fraction of what they take from society in terms of clean air, police services, road grants from the general fund and, most of all, lost time from traffic.

[A] A single driver's decision to enter a bottleneck may cost him only a few extra minutes of expensive commuting time. Totaled over thousands of drivers, congestion costs amount to millions of dollars every year on major arteries.

Also significant are subsidies to meet the physical demands of rush-hour drivers. Typically, less than a third of highway traffic occurs at or near peak times in the major direction. Yet this minority of drivers dictates the size of the road. In the Los Angeles basin, according to one academic estimate from the mid-1970s, this subsidy totals at least $250 a year for every rush-hour commuter. (Prices have about doubled since then.)

[B] University of California economists estimated 10 years ago that if Bay Area roads were priced to reflect the full costs of traffic, rush-hour charges in the central city would reach 28 cents a mile and eight cents in less dense urban zones.

Off-peak charges might amount to only a penny a mile. Yet the California gas tax today comes to only six-tenths of a cent per mile. Most transportation maladies can be traced directly to this handsome subsidy. Since individual autos pay so little at rush hour, people move farther from work in search of cheap land, causing sprawl, pollution, wasted energy--and more traffic jams. Faced with more traffic, long-distance commuters demand more highways, further burdening the public treasury. Transit agencies, public or private, too often wither in the face of such heavily subsidized competition.

Proper road pricing could work a dramatic cure. Faced with paying their fair share at rush hour, many more commuters would stagger their work hours and take carpools or mass transit, unjamming the roads. All vehicles would run faster as a result, netting most commuters a substantial cost saving in time beyond their road fees. With faster turnaround times, bus systems would make more efficient use of vehicles and drivers, and would appeal to many more riders. Air pollution would drop immediately. And, last but not least, the public could save a fortune by postponing or canceling new highway construction and by cutting back subsidies to competing transit operations.

With proper planning and explanation, there's every reason to think the public would welcome a pricing test. After all, people already pay for the roads; the proceeds of a road-use fee would replace existing taxes and finance new construction where warranted, not disappear into a black hole. Unlike general taxes, such fees would fall more fairly on drivers in proportion to their burden on society.

Just as people expect to pay for such basic services as water and electricity, they should quickly adapt to rush-hour road fees. All the more so since consumers don't seem to mind paying differential rates depending on the time (hour, day or season) for movies, air flights, telephone calls, fruits and vegetables and even some public transit. People understand that when supplies of a good are limited and demand is high, prices go up. Peak-hour road space is no different.

All very fine in theory, but how would road pricing work in practice? No one wants to fumble for change and stop every mile at a toll booth. Fortunately, better alternatives have already been proven in numerous applications at home and abroad.

In 1975, Singapore introduced a simple auto-license sticker to charge low-occupancy cars entering the crowded city center at peak morning hours. At $1.40 a day or $30 a month, it immediately cut congestion 4007o. A small force of traffic police located at entry points keeps drivers honest at low cost to the city.

Hong Kong undertook a more elaborate Electronic Road Pricing experiment from 1983 to 1985. Cars equipped with a small, solid-state electronic license plate automatically identified themselves to sensing loops in the road connected to a central computer. The computer recorded the tolls and billed car owners each month. The system permitted total flexibility as to which roads to toll and what to charge.

The experiment showed the technology to be "quite simple and of proven reliability," in the words of one government report. Politically it failed, however; motorists came away unconvinced that their fees would be offset by lower taxes. Some also objected to having their routes monitored by computer, despite the greater intrusiveness of bank and telephone records. (Anyone wanting to give the government the slip could just take a taxi.)

Similar technologies are already in use in the U.S. In 1984 the state of Oregon installed electronic transponders at several weigh stations to speed up truck identification and clearance. Participating trucks with electronic license plates can roll over weigh-in-motion scales without ever stopping. Everyone saves time and gains information. "It's a great management tool for government, a great fleet management tool and especially useful for monitoring trucks carrying hazardous materials," notes Barbara Koos, an analyst with the Oregon Department of Transportation. The encouraging results have prompted 13 states to expand the experiment and develop standards. Beginning next year, the so-called Crescent Project will establish common electronic truck identification systems from the Washington-British Columbia border to Texas.

The New York-New Jersey Port Authority plans to affix electronic license plates that will assess tolls on about 3,000 buses passing daily through the Lincoln Tunnel. Faster and less grumpy than human toll collectors, the system "should help provide an accurate reading of vehicles, speed traffic and minimize pollution and time constraints," says Joe Klementowicz, a senior engineer on the project. "If it works, we'll be looking at other vehicle populations. There is the possibility of going to private autos."

Absence of Political Will

Electronic identifiers already serve cars in some private parking lots. And California's state transportation agency (Caltrans) is testing an electronic car ID system at the Coronado Bridge in San Diego. "My hope would be that we could use them on all nine toll bridges we operate," says Caltrans's project director Jerry Meis. "The big plus for Caltrans is fewer collectors and for motorists it's convenience."

Electronic systems should cost drivers less than $20 a "tag." These small, rugged devices are powered by a transmitter at the tolling point and should last the lifetime of the car.

Some transit authorities prefer lower-tech optical scanners of the sort used at supermarket checkout stands (only much more reliable). Drivers purchase bar-coded stickers for their rear windows--or their helmets in the case of motorcyclists. In use by the Delaware River Port Authority for 15 years, the system's reliability exceeds 99%, according to Stan Shultz, vice president of Automatic Toll Systems in Mount Vernon, N.Y.

The technology is cheap, too. For only $20 a day, toll authorities can save at least four human collectors per lane. The stickers themselves cost only about 65 cents each to produce and are nearly impossible to counterfeit.

Take your pick: The technology is here to do the job. All that stands in the way of relieving traffic misery is an absence of political will. Commuters of the world unite, you have nothing to lose but endless delays on the road.

    1. Suppose that during rush hour in California's Bay Area, each additional car on the road adds an average of five seconds to the commuting time of two hundred other people. Suppose also that people value their time at $9 per hour (¼ cent per second).
    a. What is the total dollar value of the costs that each additional driver imposes on other drivers?
    b. With no congestion tolls, will there be too few, too many, or just the right number of Bay Area drivers during rush hour? Explain briefly.
    c. If congestion tolls could easily be charged, how much should each driver pay during rush hour in order to have the efficient number of cars on the road?

2. Consider passage A: Is this "cost" an externality to the driver? If so, why? If not, then what, if any, externality is associated with rush-hour commuting?

3. Consider passage B: Why should traffic charges vary so much over time?

4. Several years ago it may have been reasonable to simply put up with peak-period congestion, but judging from the article, today we should do something about it. What innovations make it more feasible now than in the past to charge drivers for peak-period highway use?

5. Name two or three other economic activities that are subject to periodic congestion.