Nine miles per hour. Few predictions have stood the test of time better than the R.J. Smeed's. Back in 1949, the statistical genius hypothesised that the average speed of traffic in London is fundamentally stuck at nine miles per hour. It doesn't matter whether new roads are built, whether better cars are designed, the speed ain't ever getting faster than nine miles per hour.
Smeed thought that nine miles per hour was the minimum speed a driver would tolerate. Any slower and they'd switch to public transport or stay home. Any faster and the reverse would happen. Smeed's been right all the way up to 2012, where the average traffic speed in central london was 8.98 miles per hour.
It's got worse. It's now a sluggish 7.4 miles per hour thanks to major overhauls to our road system and a boom in new uses like home deliveries which are less sensitive to road speeds.
Smeed's prediction held true for so long because he thought like an economist.
Roads are scarce resources just like oil and coal. And as history has shown (100 years since the Bolshevik revolution) when scarce resources are given away for free they're overconsumed.
Instead of willingness to pay, they're distributed by willingness to wait. Brits may be famously polite queuers, but it's Russians who've had the practice.
It means our roads are clogged by those who place the lowest value on their time.
Smeed knew this. That's why he advocated for road pricing as early as 1964. And why the Adam Smith Institute has advocated it since the early 80s.
Interestingly, the London Transport Committee has come around to the idea too.
When it works road pricing means that those who need to use the road the most spend much less time stuck in traffic.
The problem is London's first attempt at road pricing, the congestion charge, was badly implemented. It's set so low that it barely affects traffic flow, and it's set at a uniform rate, so it fails to push drivers toward off-peak hours and routes. Motorists see it as just another tax (and they already pay a lot of tax).
As a result, we've got a congestion charge that's way too low and way too inflexible. Not a proper system of road pricing.
But, I bet motorists wouldn't hate paying more if they'd seen the alternative – Singapore.
Singapore squeezes five and a half milion people into a city half as big as London. Yet, traffic isn't a problem. The average speed in rush hour is nearly 20 miles per hour.
Imagine that. If you lived in Singapore rather than London, you'd spend half as much time stuck in traffic. That's great for motorists but it's also great for pedestrians; cars pollute more in traffic than in free flow.
Singapore's had congestion charging since the 70s and brought in the world's first digital congestion charge in '98. And they've set their congestion charge based on consumer reality not what's politically easy (it helps that there is no functional opposition in Singapore).
It means they can have variable pricing. Encouraging efficient use in rush hour and giving less-well off drivers bargain prices the rest of time. The tech developed in Singapore could in theory allow for Uber style surge pricing when demand increases rapidly.
That'd be useful when tube unions bring London to a standstill. Rather than traffic grinding to a halt the congestion charge could rise to keep traffic flowing smoothly.
Properly pricing roads would of course raise revenue for city hall to fritter away on vanity projects. But the beauty of the system is that it'd make things much better whether we spent the money on strike-proof driverless tubes or simply burnt it.
Indeed, with the cost of congestion an estimated £8bn a year, it seems we're already quite used to burning money.