End rush hour traffic jams.

During the Cold War, we Americans were forever pointing to television images of Soviet shoppers standing in interminable lines as a testament to the superiority of capitalism, which allocates goods and services through pricing rather than long waits. Then we would smugly jump into our cars, drive onto highways, and sit in bumper-to-bumper traffic. Marxists might call this an "internal contradiction." Give us slow service at Denny's and we demand to see the manager. Stick us with traffic jams and we're sheep.

Frustrated sheep, to be sure, and getting more frustrated every year. From 1975 to 1987, the share of interstate highways in metropolitan areas that were congested during rush hours rose from 42 percent to 63 percent, according to a study by Clifford Winston and his colleagues at the Brookings Institution. The reason is simple: More people are driving more cars longer distances. While the U.S. population grew only 15.9 percent from 1975 to 1989, the total number of cars on the road soared almost 50 percent, thanks largely to suburbanization and the rise of two-earner families. Meanwhile, the length of the average commute grew by 27 percent.

The result is that we are spending an ever-greater portion of our ever-shrinking leisure time idling in traffic. Not that we mind commuting per se. Indeed, those minutes of privacy between work and home when we can decompress, alone with our thoughts and CDs, are some of the great if unadmitted pleasures of modern life. The problem is the congestion, which turns delightful 20 minute trips over swiftmoving streets into punishing 50 minute, bumper-to-bumper ordeals. This has been good for talk radio hosts and the car phone industry. It's been awful for our nerves and terrible for the economy. A study by the Texas Transportation Institute estimates that traffic congestion costs Americans $43.2 billion annually in lost time, as measured in average wages.

Fortunately, there is a fairly straightforward way to reduce or even eliminate clogged traffic: Make drivers pay tolls during rush hours. After all, space on roads during peak periods is a valuable commodity. By giving it away for free, the government naturally creates the same chaos that would result if your favorite restaurant started handing out free meals.

The concept of using tolls to reduce traffic congestion (as opposed to financing highway construction, which is what traditional toll roads do) is hardly new. For decades the idea has been advocated by economists and traffic experts, most recently by Brookings scholar Anthony Downs in his book Stuck in Traffic. But it's never been seriously tried, mostly because the notion of charging voters for the privilege of driving to work is one for which most politicians have managed to contain their enthusiasm. As Downs points out, traffic congestion is a quintessential modern public policy dilemma, one in which the real solution is politically unacceptable while the acceptable fixes won't work.

Examples of the latter--palatable but ultimately pointless--include the most obvious approach to congestion: building more roads. It's certainly true that road building has not kept pace with auto use. From 1982 to 1988, the total number of miles driven by all vehicles in 39 large metro areas grew twice as fast as roadway capacity. But as any highway official will tell you, more and bigger roads don't necessarily unsnarl traffic. That's because of a phenomenon Downs calls "triple convergence": Any increase in the carrying capacity of a given road lures commuters away from other roads, or from other travel times, or from mass transit. Road improvements do tend to shorten rush periods. But they also encourage more development, and hence more traffic.

Another approach that won't help much is beefing up mass transit. It's not just that Americans have an aversion to riding to work on trains and buses. It's that the economy is now predominantly suburban, and suburban factories and office parks are dispersed over such a wide geographic area that even devoted mass transit experts can't figure out how to link them up conveniently. Only 3.3 percent of suburban workers currently commute by public transit. Doubling that percentage would be a Herculean achievement. Yet even that would at best reduce congestion by 3.7 percent, a gain that would be offset in one to three years in a typical burgeoning suburb simply by the continuing growth in auto usage by other commuters.

A third approach, favored by both the Bush and Clinton administrations, is something called the intelligent vehicle highway system (IVHS). This involves electronic sensors embedded in roadways that send information via centralized computers to TV screens on the dashboards of individual cars, telling drivers which routes are least busy. The Department of Transportation is currently spending hundreds of millions of dollars developing IVHS. Unfortunately, it won't solve the congestion problem. Millions of drivers already enter clogged expressways every day knowing full well what they're getting into. Telling them which routes are marginally clearer will have the same marginal effect as improving roads: Commuters will switch their routes or schedules and descend on the openings like shoppers at a blue light special.

The one approach that probably would work is based on the premise that drivers who insist on taking up limited road space during rush hours ought to pay a premium for doing so. The concept, called "peak period pricing," would turn all major thorough fares in a metro area into toll roads. The tolls would be zero during most of the day, then rise during the AM and PM rush hours to levels sufficient to discourage just enough drivers to keep traffic flowing reasonably swiftly. Commuters could avoid the toll by shifting their travel times to non-peak hours or by taking mass transit where available. Or they could reduce their out-of pocket costs and still travel during peak times by car pooling.

One reason why no one has ever tried peak period pricing is that the technology to make it work smoothly has only recently been developed. Old-fashioned toll booths actually hinder traffic as drivers queue up to pay toll-takers or toss change into baskets. But now the same electronic advances that are behind IVHS can make it possible to charge tolls without forcing cars to slow down. Roadside sensors can read signals from small transponders placed behind the license plates of individual cars. Centralized computers then tally the charges and mail drivers monthly bills or deduct sums from prepaid accounts. The whole apparatus could be quickly paid for out of revenues from the tolls.

In addition to the delight of driving from one side of town to the other with your foot on the accelerator more than on the brake, peak period pricing offers profound advantages for society as a whole. An administratively and technologically feasible congestion pricing system, if applied nationwide, would save about half the $43 billion in travel time Americans now lose, estimates economist and traffic expert Kenneth Small of the University of California at Irvine. Mass transit systems would see substantial increases in ridership. Auto emissions would drop and air quality would improve as more people car pooled. Corporations would promote telecommuting and flexible work schedules even more seriously than they do now. Governments could save tens of billions of dollars in no longer needed road expansion projects, and with price signals to guide them, politicians and highway engineers would be more sure that those roads they do build are actually needed. There would be plenty of funds available for new infrastructure, too: A nationwide congestion toll system would generate an estimated $58 billion in new revenues. Finally, the regional agency that would have to be created to administer the toll system would move society a step closer to a badly needed reform: metropolitan government.

The idea of using prices to regulate traffic could be expanded, along with the benefits. We could charge airlines stiffer fees for peak period landing rights, thereby reducing both flight delays and the need to build lots of new air ports. Countless billions in road repairs could be saved, and the rail and barge industries strengthened, if trucks were charged much higher fees to reflect the much greater damage they inflict on roadways. Road congestion could be further reduced by ending the tax break employers now get for providing their workers with free parking. Several studies have found that when firms stopped providing free parking, the percentage of employees who drove to work solo dropped from 66 to 39 percent.

There is only one inarguable and unavoidable downside to peak period pricing: Most people would hate it, at least at first. The ability to drive anywhere, anytime, for free, is taken by Americans to be something like an inalienable right. The very idea of rush hour tolls would outrage the public and create golden opportunities for ideologues--especially talk radio hosts, who could be counted on to oppose any reform that threatens to reduce their drivetime . Conservatives would vilify the idea as an attempt by an expansionist state to foist surreptitious taxes on the citizenry. (Conservatives like tolls, but only on privately owned roads). Liberals will worry about the working poor, who can already barely afford the gas and upkeep on the cars they need to get to suburban jobs. Fortunately, the Regressivity Argument can be easily met: Government could simply credit electronically the toll accounts of low-wage workers. The Leviathan Argument could be handled by using at least some of the billions in toll revenues for tax relief.

This is not to say that there wouldn't be technical difficulties and unintended consequences. For instance, traffic engineers would have to figure out how to keep non-metered side streets from being swamped with drivers looking to avoid tolls. There are other questions as well. Will peak period pricing help or hurt downtowns? Will it accelerate or slow suburban sprawl? No one knows.

But there's a good way to find out: Try the concept in one congested metropolitan area and see what happens. It should hardly be politically difficult to dismantle the tolls if too many problems arise. My guess, though, is that the public will eventually grow to appreciate and support a system that allows them to spend less time with Rush Limbaugh and more time with their kids.

Paul Glastris, an editor of The Washington Monthly in 1986 and 1987, is a correspondence with U.S. news & World Report.