Stuck in traffic

Our rush hours rank with the world’s worst. Andrew Coyne has the solution.


For one in four Canadians, the two-way commute takes more than 90 minutes. (Nabeel Syed)

Day breaks over Canada, and across the country, the morning commuter rises, dresses, hops into his car and is transformed into . . . traffic. Immobilizing, enervating, infuriating traffic, glaciers of metal improbably forcing their way down the nation’s roads each morning, only to have to force their way back up the same roads later in the day.

In Halifax, drivers seethe as they inch through the Armdale Rotary. In Montreal, it’s the seemingly hours-long grind along the infamous Autoroute Décarie. Toronto commuters visibly age waiting for something to move on the “Don Valley Parking Lot.” Calgarians have ample time each day to regret taking Deerfoot Trail, while in the Lower Mainland of B.C., drivers debate which is worse: the bottleneck on the Port Mann bridge or the eternal stretches of Highway 1 on either side of it.

We’re not imagining things: traffic really is getting worse. Statistics Canada reports the average time spent commuting to and from work nationwide increased from 54 minutes in 1992 to 63 minutes in 2005. In a year, that adds up to about 32 working days spent sitting in traffic (five more than in 1992). And that’s the average. In Calgary, it’s 66 minutes; in Vancouver, 67; in Toronto and Montreal, it’s now up to nearly 80 minutes a day. For one in four Canadians, the two-way commute takes more than 90 minutes.

In part that’s because people are travelling further to work: commute distances have increased 10 per cent in a decade. But it’s also because everyone’s moving slower: average rush-hour traffic speeds in Toronto, for example, declined by 24 per cent between 1986 and 2006. The result is to make these trips much longer than they need to be: as much as 37 minutes—nearly half—of the average Torontonian’s daily commute is due to traffic delays. In a year, that’s an extra 18 days in the car.

Indeed, for sheer mind-numbing, soul-destroying aggravation, traffic in our largest cities can compete with any in the developed world. A Toronto Board of Trade report earlier this year looked at commuting times in 19 major European and North American cities. Toronto’s ranking? Dead last: worse than New York or London, worse than Los Angeles. But other Canadian cities were scarcely better. Montreal was 18th, Vancouver 14th, Calgary 13th, Halifax 10th.

It’s not just the commute. There is nearly as much traffic at lunchtime today as there was at rush hour a generation ago. Not only are there more cars and trucks on the road—21.4 million registered vehicles, up from 16.6 million in 1992—but we’re using them for more things: driving the kids to sports, where once they would have walked. Total daily trips in the Greater Toronto and Hamilton area rose by 56 per cent between 1986 and 2006.

Traffic is slowly strangling our cities. It’s the time wasted in traffic that could have been put to more productive use. It’s the late deliveries, the missed appointments, and the margin of error needed to cover the risks of either. It’s the extra repair costs from all those additional fender-benders. It’s the higher fuel consumption and consequent higher emissions to which stop-and-go traffic gives rise, to say nothing of the added wear and tear on roads, and tires, and engines—and heart muscles: being in heavy traffic triples your risk of a heart attack within an hour, according to German researchers. It’s the measurable drop in property values in areas overtaken by the traffic blight. It’s the noise, and smell, and general unsightliness. And much more besides.

Add it up and the costs are massive, and growing. A 2006 Transport Canada study put the cost of congestion nationwide, taking everyday and “non-recurring” congestion (accidents, road work and so on) together, at as much as $6.7 billion. (Interestingly, measured in congestion costs per vehicle-kilometre, Vancouver can lay claim to having the worst traffic in the country: see chart.) Yet even this is almost certainly an underestimate. The figures are in 2000 dollars, for starters, and traffic has appreciably worsened since the early years of the decade, when the study was conducted. Costs were estimated only in the nine largest urban areas, only at rush hour, only for cars (not trucks or buses), and only included the drivers’ wasted time and excess fuel consumption (and related greenhouse gas emissions).

A more comprehensive estimate, conducted in 2008 for Metrolinx, the agency responsible for transportation in the Greater Toronto and Hamilton Area, put the annual cost of the congested state of the region’s roads at $6 billion, when knock-on costs to the surrounding economy are included. That suggests annual congestion costs for the country as a whole would today approach $15 billion, nearly one per cent of GDP. Now factor in the rapid growth in population that Canada’s major cities are expected to undergo in coming decades. Something’s got to give.

And yet, nothing ever does. Though the nation’s roads and highways get more congested with each passing year, municipal and provincial governments persist in the same approaches that got us where we are today, which is to say, stuck in traffic. A City of Toronto newsletter, after happily running through some of the city’s many “traffic demand management” programs, ends by cautioning residents not to expect them to work. “We can’t solve the issue of congestion,” it says, “but we are trying to manage it better.” (Instead, readers are urged to see the “positive side” of congestion, as “the sign of a vibrant city.”)

But we can solve it. That our cities have failed to do so is not for lack of proven alternatives, but in wilful defiance of one in particular, a solution that not only has an impressive expert consensus in support of it but is already having notable success in other cities around the world. There’s even a working model of it in place right outside Toronto.

We do not have to suffer this daily indignity, in other words. It is not natural or inevitable that urban traffic should move with the speed of industrial sludge. It’s not often true of other social problems, but when it comes to traffic, there really is an Answer.


LET’S DEAL first of all with some of the more popular non-answers. Over the years a lot of utopian hopes have been invested variously in telecommuting, or staggered work hours, or car-pooling as the way to lessen the crush at rush hour. But it turns out, as the research shows, people like working downtown, almost as much as they like living in the suburbs, and they like doing so at the same time as other people. Moreover, the vast majority of them like the convenience and independence of driving there themselves: measures to encourage car-pooling such as high-occupancy vehicle (HOV) lanes have had limited impact.

So even in the face of commute times of an hour or more, people vote with their steering wheels. Today’s utopians dream of solving the traffic problem by mass defections from the car to public transit, or even bicycles. But it would have to be massive indeed. Nationwide, more than 85 per cent of Canadians continue to get to work by car, a figure that has not changed in two decades.

The reason is simple: it’s quicker by car. As bad as the commute is for drivers, it’s much worse for public transit users: 106 minutes, versus 63 minutes by car. Granted, part of the reason it takes so long to get anywhere by transit is because of all the cars blocking the way. But you’d have to persuade an awful lot of those drivers to give up the comfort and convenience of their cars to put much of a dent in that. And they’d still take longer to get to work even then.

Other commonly proposed solutions would be of similarly marginal benefit. You can better synchronize traffic lights. But whatever savings in time that confers on traffic going north-south is only time added to traffic going east-west. European-style roundabouts, besides being more romantic than static four-way intersections, also make for better traffic flow. But we’re not about to tear out every street corner. And in any case, the problem is more fundamental than that: Europe’s traffic is as bad or worse as North America’s.

More and more cars trying to get through the same narrow passages is an obvious recipe for congestion. Of course, one way or another, road space will be allocated; at present, it’s rationed by time. Which is perverse, when you think about it. The people who are prepared to “pay” the most to use the roads under this system are the ones who need them least: those who place so little value on their time that they are willing to spend years of their lives, literally, sitting in traffic. Alas that leaves just-in-time delivery trucks and other, more time-sensitive travellers—ambulances, fire trucks, mothers with kids in daycare—stuck in the same jams they are.

It also means that the other favourite political remedy—build more roads—is no more the answer than public transit. Countless empirical studies have shown: add more road space, and traffic simply expands to fill it. True, at first the extra lanes or new highway do reduce congestion and shorten travel times. But reducing the “price” of driving in this way only stimulates the demand. Very soon you find more cars on the road, taking more trips. The same “induced traffic” phenomenon, by the way, can be seen in reverse: knock out a road, and traffic doesn’t drop by nearly as much as the reduction in capacity would imply.

That does, however, point us in the direction of the real answer. Because it suggests that traffic levels are not, as commonly believed, a given, as if people simply had no alternative but to drive a fixed distance every day. If they have to, they can and do cut back in any number of ways in the short term, and even more in the longer term. These include some of the ways listed above: taking transit, catching a ride with a friend, walking, or simply cutting out needless trips. They can travel, part of the time at least, at off hours, or on less-congested roads. Given enough time, they can even alter living arrangements, living closer to their place of work, shopping closer to home. But they won’t do any of these until they have an incentive to do so: until we find a more rational means of allocating road space than time, one that actually encourages people to economize their use of the roads rather than simply scramble to be at the front of the line.

As it happens, we have such a mechanism. It’s the one we use to allocate resources everywhere else in the economy: prices. Or if you prefer, tolls.

THE IDEA is hardly new. Modern thinking on tolls, or “road pricing,” dates back to the early 20th century economist Alfred Pigou. As long ago as 1964, the Smeed Report in the U.K. was making practical proposals for its implementation. Closer to home, there have been any number of reports, studies, and expert panels taking up the issue of late, all making the same recommendation: charge people to use the roads. A short list from the last couple of years would include the C. D. Howe Institute, the Organisation for Economic Co-operation and Development (OECD), the Ontario environment commissioner, the Toronto Board of Trade, the Toronto City Summit Alliance, British Columbia’s TransLink (the Lower Mainland equivalent of Metrolinx), the City of Montreal and the City of Toronto.

Nor is the analysis terribly complicated. In essence, roads are an example of the “tragedy of the commons.” The failure to charge for them leads to their overuse, the same way an open pasture will soon be nibbled bare, as every farmer races to be the first to bring his sheep to graze for fear of being crowded out by the rest. It should not be surprising, then, to see the consequences: not only congestion, but its close cousin, sprawl. It seems odd that the more area a city covers, the more congested it becomes. But the reason for each is the same. We have made travelling by car artificially cheap in terms of money, and artificially costly in terms of time.

But wait a minute, you say: I paid for those roads already, via the gas tax. Isn’t that a kind of user fee? Not so fast. You may have paid for the roads. What you haven’t paid for is you. Every time you use the road, you take up space that might be occupied by another car. And so far as you do, you impede the driver of that car from getting where he wants to go, and doing what he would rather be doing there. Since time is money, you impose a cost on him—as does he on you.

Of course, any one driver only slows up the mass to a small extent, but add up the “external” costs of every driver on the road—and it comes to a great deal, as we’ve seen. The obstruction need not be overt, as in a stalled car or accident. Just the presence of other drivers is enough to cause drivers to slow slightly, or to miscalculate, get too close, and have to slam on the brakes. The ripple effect of any one car doing so is then transmitted back through the line, exponentially worsening the initial disturbance. That’s how we get traffic jams.

The task, then, is to make those “external” costs apparent to each driver. Raise the price of using the roads, and people will reduce their demand for it, just as they do for most other things. That makes driving more expensive, at least at first. But with less congestion, other costs fall. Not only are travel times reduced, but so are all those other ills of congestion, from accidents to pollution.

So: jack up the gas tax, then? No, because congestion is a phenomenon, not so much of roads in general, but of particular roads, at particular times. Fuel taxes may be a good rough proxy for distance travelled, but they take no account of which roads you travel on, or at what time of day. For that you need a more precisely targeted “congestion price.”

Does that mean zero congestion should be the goal? No. While there’s an inverse relationship between traffic density—how closely packed the cars are—and speed, what matters is the product of the two, traffic volume: how many cars pass through a given stretch of road in a given time interval. So a certain amount more density is worth a certain amount less speed. How to judge the trade-off? How about leaving it to drivers themselves? Up to a point, they’ll be willing to pay to reduce congestion. But past that point, the additional savings in time and other costs won’t be worth it to them. It’s the excess congestion we want to eliminate—the part motorists would be willing to pay to avoid.


THAT SOUNDS THEORECTICAL. But in fact there are plenty of real-world examples. Anyone who has driven the toll-funded autoroutes in France will know what wonders they are: fast, glass-smooth, with pleasant rest stops every few kilometres. You enter at certain restricted access points, and pay as you leave, depending on your distance travelled. They’re not cheap: about $65 to drive from Paris to Marseille. But there’s another point. It turns out drivers are willing to pay for good roads: more, perhaps, than they might be allocated around the cabinet table, where they must compete against health care and—well, against health care. And the toll does its job of rationing demand. You find yourself weighing the options, calculating: am I in that much of a hurry today? Or can I afford to take the smaller, slower, “free” road?

On some U.S. highways, you get both options in one. Drivers on Los Angeles’s notoriously congested Riverside Freeway can take the Express Lanes, a privately built freeway-inside-a-freeway. The price varies by the hour, depending on volume, with a view to keeping traffic flowing at the speed limit: from $1.30 late at night to as much as $10.25 during “super peak hours.” (That’s not just a targeted speed, by the way: you get your money back if it’s any slower.) It puts through twice as many cars per hour as the free lanes, at four times the speed. On Interstate 394 near Minneapolis, drivers can likewise pay to join the carpoolers in the high-occupancy toll (HOT) lane, a variant of the HOV concept. Again, the price fluctuates with demand; roadside signs advise drivers of the going rate. Since opening in 2005, it has increased average speeds by 15 per cent.

But why cite foreign examples? Just north of Toronto, running roughly parallel to Highway 401, is the privately owned Highway 407 Express Toll Route. Built in 1997, it is among the busiest toll roads in the world, with more than 360,000 trips taken along its 108-km route on a typical workday. There are no toll-booths: drivers can enter or exit where they like, their presence recorded by overhead “gantries,” or sensors. Rather, tolls are collected and administered electronically, via on-board transponders (for regular customers) or by snapping a photo of the driver’s licence plate and billing them by mail: between 18 and 42 cents a kilometre, depending on the type of vehicle and the time of day.

Because it depends on repeat business for the bulk of its revenues, the 407’s owners have a strong incentive to keep the traffic flowing as smoothly as they can. Tolls are only part of that. The company offers free 24-hour roadside assistance, for example, including boost, tire change, gasoline, and a tow truck if needed. The highway is noticeably well-maintained and pothole-free, especially by comparison to its often impassable neighbour.

Of course, who knows what the 401 would look like had the government simply put a toll on it? Maybe if it had, it would not have been necessary to build the 407 at all. While cash-constrained governments would be well advised to make any new road construction toll-funded—as a test of demand—they should first test whether better use could be made of existing roads.

Still, just tolling the surrounding highways won’t be enough to clear Canada’s major cities of congestion. Rather, we need to learn from those cities around the world who are applying tolls within city limits. Probably the most famous of these is London’s congestion charge. Introduced in 2003, it is an example of a “cordon” toll, collected at a number of entry points ringing the city centre. The price is steep: $15.50 on weekdays (up from $7), charged to drivers using licence plate recognition. The plan has not been without controversy, notably over an abortive attempt to expand the toll zone, but there is no doubt it has succeeded in its stated aims: a 20-30 per cent reduction in traffic flows across the cordon in the first year. So immediate and striking was its success, the mayor who introduced the charge, Ken Livingstone, was re-elected the year after.

A similar story has unfolded in Stockholm. Beginning in January 2006 as a seven-month trial run, weekday visitors were charged a fee, varying according to traffic volumes: from $1.50 in off-peak hours, to twice that much at peak. As in London, traffic flow into the city centre was reduced by more than 20 per cent. Transit use soared; there were fewer accidents; vehicle emissions declined. Small wonder that in a referendum some months after the trial ended, Stockholm residents voted to make the arrangement permanent.

To be sure, tolls have aroused much public opposition elsewhere—referendums in Edinburgh, Manchester, and several Swedish cities failed—but always in advance of their introduction. Where tolls have actually been implemented (Milan, Oslo, and Melbourne are other examples), they have never been withdrawn.

Still, there’s a flaw in these schemes. After initial sharp reductions in congestion, both Stockholm and London saw some erosion of these gains in later years: though many fewer vehicles were entering the city centre, as before, the reduction in traffic within the cordon was much less. Why? Much as if the city had added capacity, drivers responded to the easier traffic conditions by . . . driving more. So perhaps we need something still more ambitious—something like Singapore’s “electronic road pricing” scheme.

With Singapore’s long experience of London-style cordon tolls—their own scheme, implemented in 1975, was the world’s first—the city state took things a step further in 1998. Not only are tolls collected at entry points, but also along major arterial roads. Every vehicle on these roads must carry a card on its dashboard, much like a prepaid phone card, capable of being read by Highway 407-style sensors: the value of the toll is deducted from the card automatically, with higher rates applying as traffic volumes increase. Drivers therefore have an incentive to choose less heavily travelled roads, but without a backfill of traffic flooding in to take their place: prices see to that.

Indeed, the system works so well, the question arises: why not toll . . . every road? Obviously this couldn’t be done with toll booths, or even gantries. But with satellite tracking technology, familiar to anyone who uses a GPS, it should be possible to apply the Singapore model comprehensively. Tolls would no longer be discreet events, but more like your phone bill: the price you paid to use the road system, much as you pay to use the telephone network. The tolls would vary dynamically, according to the time of day, the distance travelled, the type of vehicle and so on. Satellites, moreover, could be used to update drivers on the prices of different roads as they came up; route-planning software could be used to predict the costs of alternative routes.

If that sounds far-fetched, you should know that the British government under Tony Blair came within a hair’s breadth of implementing just such a scheme. In a white paper published in July 2004, it proposed a rate schedule ranging from a few pence, for weekend drives in the country, to more than $1.50 a mile, for rush-hour traffic on the ring road around London. It was calculated the plan could reduce the amount of time lost in traffic jams by nearly 50 per cent. The aim, Blair declared in 2006, was to introduce “a national road-user charging scheme . . . within the next decade.”

Alas, the plan was later abandoned by Gordon Brown in the face of popular opposition. But the idea is far from dead. The Netherlands was all set to introduce a similar scheme this year, before a member party in its governing coalition got cold feet. Oregon has experimented with it. Trucks in Austria and Germany already pay tolls this way. Indeed, some of its strongest proponents are to be found here in Canada. Toronto-based Skymeter Corp. is actively marketing the technology, while policy gadfly Lawrence Solomon, founder of the free-market environmental group Energy Probe (disclosure: I am an unpaid director of Energy Probe), holds several international patents on it. The Toronto City Summit Alliance treated the idea seriously in its recent report. A demonstration project in a major Canadian city might be just the thing to launch the technology worldwide.

There are obvious practical obstacles to implementing such a plan, though none that seem insurmountable. Privacy is a common objection, but similar concerns do not seem to have prevented millions of people from entrusting the records of their most intimate conversations to the phone company: it is surely possible to be as discreet with the usage of their car. There are simple technological fixes, for example, converting data on a car’s location to the corresponding price before it ever leaves the transponder. Those for whom it remained an issue could prepay, again on the cellphone model.

The problem of enforcement, likewise, is more apparent than real. You would be required to install a transponder as a condition of licence, just as you are obliged to have a working odometer, tail lights etc. Spot checks would be easy enough to conduct.


WHAT ABOUT the more fundamental objections to road pricing? Two in particular come to mind. The first, that tolls would be unfair to the poor, is perhaps the more easily discarded. The very poor, of course, would not pay the tolls, as they do not typically have cars to drive. The rest could be compensated in cash, similar to the GST tax credits, rather than giving everyone, rich or poor, a free ride. And of course, the poor benefit as much as anyone from clearer streets and faster travel times, not least as transit users.

To the second, that tolls would become a cash cow for governments, the simple answer is that any revenues from tolls can and should be used to lower taxes: perhaps even the gas tax. To be persuasive, the offset would have to be guaranteed, immediate, and 100 per cent: voters are rightly skeptical of any such promised trade-offs.

Indeed, it may even be necessary to go so far as a plan put forward recently by the Social Market Foundation, a British think tank. It proposed putting ownership of the road system in a public trust, at arm’s length from the government. At the end of the year, all toll revenues would be distributed to every member of the public—the trust’s shareholders. Depending on how much you drove, you might even make a profit on the deal.

Which means discarding one of the most common arguments made for tolls: that the revenues could be used to finance public transit. For starters, this is unnecessary: the very act of tolling roads would, by itself, make public transit more competitive, since the per-person cost of the toll would be much less for buses than for cars (and none at all for subways and surface rail). Moreover, as the economist Robin Lindsey explains in a study for the C. D. Howe Institute, “transit vehicles speed up when tolls are imposed, because there are fewer cars on the road. This attracts more travellers to transit. In response, transit operators improve service by adding routes and increasing frequency. Due to economies of scale in transit operations, the cost per passenger falls, perhaps allowing the operator to lower fares. Ridership increases further, and so on.”

If getting more people to use transit is your aim, moreover, subsidies are the last thing you should want. The biggest factor in people’s decision whether to use transit is not the fares, but rather the speed, comfort and convenience relative to other options: that is, the passenger experience. And the surest means of forcing transit operators to pay more attention to the passenger experience is if their livelihoods depend on it. The greater the share of revenues paid for by passengers themselves, the more operators are likely to be lying awake at night thinking up ways to put bums in the seats; subsidies simply insulate them from that concern.

The nub of the argument, whether we are talking about cars, or buses, or tennis rackets, is this: people make better decisions when they know what things cost. Right now the true cost of using the roads is hidden, leading people to drive more and in different ways than they would if they were better informed.

Even a modest road-pricing scheme would be a start: traffic jams wouldn’t be entirely a thing of the past, but they would be a lot less common. And the more comprehensive the plan, the greater the payoff: shorter travel times. Lower fuel costs. Fewer accidents. Less noise and pollution. Higher productivity. Road pricing would make us richer, healthier, saner. If London, Stockholm and other cities can do it, why can’t we? Why, other than because it would be new, and because we would be paying for something we were used to getting for free.

Only it isn’t free now. It’s hideously expensive. There ain’t no such thing as a free lunch, and as any commuter can tell you, there sure ain’t no such thing as a free road.

Looking for more?

Get the Best of Maclean's sent straight to your inbox. Sign up for news, commentary and analysis.
  • By signing up, you agree to our terms of use and privacy policy. You may unsubscribe at any time.