Using Game Theory to Make City Streets Safer

Part of this problem is the over-ruling of our unconscious assumptions (sometimes described as system 1 thinking). Not only is driving mentally taxing, most of us fail to see that as drivers, we are part of the problem. Take, for example, freeway congestion. A 2006 report from Canada’s transportation department reads: “Motorists do not perceive themselves as a cause of congestion; motorists perceive themselves as the victims of congestion.”⁸

In game theory, Braess’ paradox explains this phenomenon perfectly: As a new highway or lane is unveiled, people are drawn to the promise of quicker commutes, and inevitably the highway becomes even more congested.³

So what can we do about this?

In uncontrolled intersections, like traffic circles (or roundabouts), drivers are forced to focus on the players of the game: drivers entering, exiting, and continuing onward, as well as bikes. In contrast, at a light-controlled intersection, you are probably focusing most of your attention on whether you will make the green light in front of you and less on what is happening on the street. Traffic circles are commonplace in Europe for this very reason: they reduce the number of conflict points and naturally force a reduced speed.¹³

Based on this, we can explain how unmarked streets therefore become safer: They make us more conscious drivers and they force us into cooperation by adhering to fewer rules. In Bohmte, one of the seven test cities in the original experiment, they removed all signs and markings and replaced them with just two rules: A maximum speed of 30 km/h and obligation to yield to anyone on the right.⁵ Ipswich, another test city, reversed the right-of-way rule on a popular high street so that pedestrians had right of way.⁷ This makes the unconscious, conscious.

Streets with no traffic signs can be safer because they change the rules of the game. Instead of assuming the correct behavior of other drivers or pedestrians, we are required to drive with greater attention to the greater good. In other words, we are forced to cooperate.

Back in Bohmte, following the first four weeks of the new design, no accidents were reported. This was noteworthy since the town had previously had at least one major accident per week.⁵ Furthermore, the fluidity of the traffic increased and start-stop traffic was reduced.¹²

Behavioral science has been applied in various aspects of city design, for example with transportation or commuting habits. This is relevant for our current situation because one result of the COVID-19 pandemic is that streets across the world have been temporarily transformed with new pop-up bike lanes that share the road with vehicles.⁹ Citizens and officials have voiced concerns how the temporary measure could increase traffic congestion once the pandemic is over and may decrease safety for cyclists.^16,19

One solution might be to paradoxically introduce more chaos and less order into our intersections, removing dedicated areas for bikes and letting cyclists mingle with vehicular traffic. While this may run counter to what cyclists perceive,¹ certain traffic-heavy intersections may benefit from the conversion into a so-called “shared space.”

To be clear, the research shows us that the safest streets for cyclists are still segregated bike lanes.¹ However, intersections remain one of the most dangerous areas for riding a bike, and observational data since the coronavirus shutdown has shown that in some cities, pop-up bikes lanes have not resulted in fewer cyclist deaths, but more.^9,11 For that reason, it is worth considering how and where we can bring the concept of shared spaces to work for our intersections, particularly the role of traffic circles.

Of course it is worth noting that not every intersection or street would be improved by removing all rules. We can look at the ongoing discussions about the true comparative safety of Germany’s Autobahns to see why reviewing situational factors is always necessary.¹⁸ However, playing by these paradoxical rules might help us in the present, and could come to decide what we do with pop-up bike lanes in the future.

The computational load of driving with many other agents on the streets (drivers, cyclists, pedestrians, stray animals, flyaway sofas, etc.) taxes our brains to an extent that sometimes our autopilot mode can fail us and result in an accident. “Shared space” roads are an example of paradoxical urban design that can mitigate this behavior. It prescribes removing most signage and barriers between different actors.

The result is that drivers will have less tunnel vision on the traffic lights ahead, and instead scan their entire driving environment, seeking eye contact with other drivers, cyclists, and pedestrians. Data from seven test cities in Europe from over a decade ago show that shared space intersections can be a safer alternative for pedestrians and cyclists.

The time to revisit this decade-old experiment is right now, as the global pandemic has forced us into different commuting patterns. Shared spaces offer a potential solution to keep intersections safe once our commute regains some of the normalcy of pre-COVID times.