In 2002, central London was, to all intents and puoses, a perpetual traffic jam. On a typical day, a quarter of a million vehicles would drive into the eight square miles of central London, ready to do battle with a million commuters who used public transportation. In place of long, wide boulevards, London has tightly packed, narrow, winding streets, which kept the average speed of traffic below ten miles an hour. On a bad day, it was more like three miles an hour. You can walk faster than that and not even break a sweat.
Traffic was so bad, in fact, that it turned the mayor of London, Ken Livingstone, from an avowed socialist into the advocate of a plan that warmed the hearts of capitalist economists everywhere. In February of 2003, London started charging people to drive into the centre. If you wanted to enter central London between 7 AM and 6:30 PM, you now had to pay £5. If you neglected to pay, and one of the 230 cameras the city had installed recorded your license plate, you got stuck with an £80 fine. In theory, the plan was supposed to raise £180 million a year for the city to invest in public transportation, and to cut traffic congestion by 20 percent.
The principle behind the London plan was a simple one: when someone drives into the city and makes traffic worse, he inflicts costs on everyone else that he never pays for. When you’re that driver and you’re sitting in bumper-to-bumper traffic while toddlers speed by you on the sidewalk, it feels like you’ve paid more than enough. But the mathematics of congestion suggest that you haven’t. The toll is an attempt to collect the bill.
“Congestion pricing” has been around as an idea since the 1920s, but its most important advocate was the Nobel Prize-winning economist William Vickrey. For Vickrey, road space was like any other scarce resource: if you wanted to allocate it wisely, you needed some way to make the costs and benefits of people’s decisions obvious to them. Because, say, the main road into the city is free, everyone chooses to drive on that road during rush hour even when it would be better for just about everybody if some of them drove earlier or later, some took public transportation instead, and some worked from home. If that same road had a toll on it, different people would make different choices, because they would have different answers to the question, “How much is this trip really worth to me?” And so, instead of everyone ending up on the same road at 6:30 PM, they’d leave work earlier or later, or take the train, or telecommute.
It’s a nice idea in theory, but putting it into practice has always required a very hard sell. Livingstone had to fight off massive lobbying efforts in opposition to his plan for London. In thç United States, meanwhile, congestion pricing has always been a non- starter. Americans don’t like paying highway and bridge tolls, but they hate the idea of having to pay more money to drive during rush hour. Most people feel as if they have no choice about when or how they commute, and the thought of the wealthy paying to zip along empty roads while everyone else takes the long way around grates. As a result, we’d rather suffer in traffic than allow some to pay for freedom. The authors of a study of a failed attempt to introduce congestion pricing on San Francisco’s Bay Bridge, for instance, concluded that both voters and politicians need to be convinced there are literally no other alternatives before they’ll accept a Vickrey scheme. There are a few exceptions to this rule, most notably New York City, where it costs more to use certain bridges and tunnels during rush hour. But there are only a few.
Oddly, we’ve happily adjusted to something like congestion pricing in other parts of our life. Long-distance calls are more expensive during the day, drinks are cheaper during happy hour, and it costs more to go to Las Vegas on the weekend than during the week. (And don’t forget the early-bird special, either.) All these are cases of price responding to demand—when demand is high, the price goes up, and when it’s low, the price goes down. But when it comes to driving, Americans seem to prefer it when there’s no price—at least in money terms—at all.
It’s not surprising, then, that the one place in the world that’s made an art out of congestion pricing is the antithesis of America in cultural terms, namely Singapore. Blessed with not having to worry about angry drivers’ groups or disgruntled voters, Singapore’s government put congestion pricing in place for the first time in 1975. The initial version of the plan looked a lot like London’s more recent scheme: you had to pay a toll if you wanted to get into the country’s central business district (CBD) during rush hour. As time went on, the plan expanded, until you had to pay if you wanted to get into the CBD at any time during the day. But the most important changes have been technological. Once upon a time, the system was enforced by meter maids who recorded the license plate numbers of rule breakers. Today, every car in Singapore has an electronic smart card attached to the dashboard, and as soon as you cross into a pay zone, you see the money disappear from your card. This has two advantages: it makes cheating impossible, and it makes the cost of your decision to drive immediately obvious to you. Singapore has also made its pricing rules more sophisticated. While there was once one price to drive during the morning rush hour, now there is “peak-within-peak” pricing (it’s half as expensive to drive between 7:30 and 8 AM as it is to drive between 8 and 9 AM), and evening pricing. Singapore even offers weekend-only cars (on which drivers get a tax break and price rebates). Not surprisingly, traffic in Singapore is much better than it is in London or New York, even though the country has more cars per mile of road than any Western country. (Of course, it is a very small country.)
The interesting thing about Singapore’s success is that for all of the country’s authoritarian ways, it has left the actual decision about whether or not to drive in the hands of the individual. One way to cure congestion, after all, would simply be to ban certain people from driving on certain days. And this, in fact, is exactly what Mexico did, albeit in an attempt to curb air pollution. If you live in Mexico City and your license plate ends with a 5 or 6, you can’t drive on Monday. (A or 8 means you’re out of luck on Tuesday, 3 or 4 on Wednesday, and so on. Everyone gets to drive on Saturday and Sunday.) But this hasn’t done much to reduce traffic, because drivers have no incentive to find alternatives to driving on the six days a week when they can drive, and because many Mexicans just bought second cars that they could use on their supposed off days. Singapore’s plan, by contrast, tells the drivers how much it’ll cost to use their cars, and then trusts that the sum of all those individual decisions about whether or not to drive will be smart.
Figuring out how much driving should cost, though, is a tough problem, and economists have spilled a lot of ink trying to solve it. One obvious challenge is t1at the wealthier you are, the easier it is to trade money for time and convenience (you’ll pay to drive into London because it’s easier than taking the tube). Poorer people can avoid the toll by not driving, but that doesn’t make them any better off than they were before. So any fair congestion-pricing plan has not only to charge tolls but also redistribute the revenue they raise. Singapore did that by building a hyper-modern mass rapid transit system, and Livingstone’s plan for London similarly involves spending hundreds of millions on public transportation. Another alternative, proposed by the traffic engineer Carlos Daganzo, is to allow people to drive for free on some days and charge them on others.
That keeps the right incentives in place, but also keeps the money- for-time crowd from dominating the highways.
An ideal pricing system would be considerably more sophisticated than London’s £5 all-day system. Vickrey, for instance, imagined a world in which traffic was governed by “responsive pricing,” so that how much you had to pay to use a road might vary depending on how heavy the traffic on that road was right then, or on the weather, or on the type of vehicle you were driving. If I-S between Sacramento and San Francisco suddenly became clogged with traffic because of a broken-down tractor trailer, it would cost you more to use it. That, presumably, would divert people to other routes, and keep the congestion from getting out of control. Today a system like this is actually technologically feasible. It is, of course, a political pipe dream, and hyper-responsivc pricing may in any case be more trouble than it’s worth. (Is it a good idea to have people carrying out complex price calculations while traveling at seventy miles an hour?) But the possibilities created by things like highways wired with traffic-detecting sensors and cars equipped with global positioning systems are endless.
Still, crude as it is, the London plan has been far more successful than most noneconomists thought it would he. Traffic has fallen by almost 20 percent, congestion has been significantly reduced, and, according to at least one study, cars are able to go 40 percent faster. (That still means they’re only going eleven miles an hour, but you take what you can get.) The biggest concern people have now is that the plan may have been too successful in curtailing driving. After all, the point of congestion pricing isn’t to stop people from driving, since from an economic perspective (and setting aside the environmental one), a highway that’s empty is hardly better than one that’s too full. The point of congestion pricing is to get people to coordinate their activities better by balancing the benefits they get from driving against the costs they inflict on everyone else. In the London case, the concerns about the traffic decline have been overblown. The roads are still full of cars.
They’re just moving more easily. More important, the flow of traffic is now a better reflection of the real value people place on driving. At least for the moment, London traffic is wiser.