16 August 2018

Capping Uber is the wrong way to cut congestion


London’s congestion problem is nothing new, but it is getting worse.

In 1949 statistician R. J. Smeed made a bold claim. He predicted that London’s average traffic speed would always be 9mph. The logic being that any faster and it’d encourage more to drive, while any slower and people would switch to public transport. It was a prediction that has done extraordinarily well over the years. In 2012 London’s average traffic speed was 8.98mph. But recently, things have gone downhill. In 2016, it fell to 7.4mph and there’s every indication that the problem has got worse.

Sadiq Khan blames Uber. He wants new powers to copy New York City and cap minicab numbers. While it is possible that Uber may have increased congestion in the capital, the Mayor of London is taking the wrong approach. Capping minicab numbers would be a massive step backwards for competition and innovation.

Khan’s plan would push up fares, increase wait times, and deny the public choice. But it’s not bad news for everyone. Incumbents hate competition and taxi drivers are no different. They will be able to charge higher fares and attract more customers, not by offering a better or faster service, but because the Mayor has hobbled the competition. No wonder they’ve campaigned for a cap for years.

Caps are blunt instruments. They only hit congestion indirectly. One unintended consequence of any planned cap is that it could encourage part-time drivers to work longer for higher fares. Another problem is that Uber competes not only with public transport but also private car ownership. Private cars tend to spend longer in traffic searching for a space to park. A recent study by parking expert Donald Shoup found that at any one time 15% of traffic will be cruising for parking (other studies have found much higher rates). One study from Arizona State University found that when Uber entered an urban area traffic congestion fell significantly.

There’s a better way to ease congestion. In a 1964 paper for the Ministry of Transport, Smeed set out how congestion pricing could prove his own prediction wrong. Almost fifty years later, London brought in the Congestion Charge. For the first time since Smeed’s initial prediction, average traffic speed in London increased to 11 mph (a 15% increase).

London’s Congestion Charge isn’t perfect. It’s set too low. It doesn’t distinguish between peak and off-peak travel. Nor does it distinguish between someone making a five-minute trip into the zone and someone spending the whole day in Central London. Most importantly, it doesn’t include taxis or minicabs.

That will change soon. Transport for London recently set out plans to apply the Congestion Charge to minicabs. This is a better approach than a blunt cap, but again it’s flawed. Once again, Khan has given into the London Taxi Driver Association (LTDA) and excluded black cabs from the charge.  TfL justify the favourable treatment on the grounds that black cabs are wheelchair-accessible, while many private-hire vehicles aren’t.

Yet, if the aim is to help the disabled then there’s a better solution. Why not include taxis in the Congestion Charge and spend the extra revenue on Londoners with serious mobility issues, for instance by increasing funding for schemes such as Taxicards?

But adding taxis and minicabs to the Congestion Charge wouldn’t be enough to seriously bring congestion down. To do that we need to replace London’s Congestion Charge with a modern road pricing system.

Other countries show us the way. In Singapore, where they have had road pricing since the 70s, the average speed of major roads in rush hour is almost 20 mph.

As Uber uses surge pricing to manage spikes in demand, so the Congestion Charge should do the same. London could nudge users into travelling at less busy times by following Stockholm and increasing the Congestion Charge at peak travel times.

Under the status quo, you pay the same rate whether you pop into the Congestion Charge zone for a few minutes or spend the whole day circling Silicon Roundabout – that should change. We should charge road users based on how long they spend driving in the zone. With past technology this would have been tricky (Smeed’s proposal decades ago of installing special meters in each car was quickly dismissed), but today smartphones are ubiquitous.

If we fixed the Congestion Charge then Uber (and its competitors) can be part of the solution to congestion, not a cause. With the costs of congestion properly priced-in Londoners might be more likely to share their rides with other passengers using UberPool or CityMapper’s SmartRide.

Sadiq Khan’s plan to cut congestion is anti-competition and anti-consumer. Thwarting markets won’t solve congestion, but using them will.

Sam Dumitriu is Head of Research at the Adam Smith Institute