In 1952, English mathematician and transport analyst John Glen Wardrop stated two principles for equilibrium route assignments for vehicles in traffic. The first principle, the principle of user equilibrium states no vehicle can reduce its travel time by changing its individual route.

In other words, for this equilibrium, travel times on all routes are the same. Whereas the second principle is for social optimal travel. This occurs when vehicles cooperate to reduce average travel time. Let’s illustrate this through an example.

Above, you see three paths from Origin to Destination: AB, AED, and CD. The travel times on each link are given in the formula in minutes, with x being the number of vehicles traveling on the link and a total of 100 vehicles in this scenario.

From setting the travel times equal on all 3 paths, we get travel time=3.75 minutes, for every driver. Thus, no driver benefits from changing their route individually, which is the user equilibrium. But is this the best outcome?

It turns out that the best outcome is when path E is not used! When drivers cooperate to not use path E, the average time is reduced to 3.5 minutes per driver! This is also referred to the Braess’ Paradox, where removing roads could actually reduce travel time.

Of course, a defector could decide to use path E. More defectors would ultimately lead to the selfish user equilibrium. But perfect coordination is clearly superior.

Capitalist vs socialist routing

In essence, user equilibrium is the outcome of a free market ideology. Every driver has same access to route information, and ultimately chooses the route that minimizes their individual travel time, or maximizes their individual benefit. The result? A sub-optimal routing strategy.

Whereas, the social optimal prioritizes global travel times. In general, it is possible that individual drivers might suffer by traveling for longer, but the social optimal strategy reduces the total travel cost, or maximizes the social benefit.

But how can a society coordinate to achieve this minimal cost? Currently, the best solutions for traffic are routing software like Google Maps, that give everyone the same information and best path.

This is the same as selfish routing, the user equilibrium. Instead, what if everyone were given different routes to maximize the social good by minimizing global travel times? You might say well I want to reach work on time, and don’t want to sacrifice for someone else. But when this game is played 261 times a year when you go to the workplace, on average you will reduce your travel time! The increasing prevalence of big data technologies will hopefully help achieve this vision.

In the near future, it might be possible to tightly coordinate routing and streamline transportation. This can be extremely important during emergencies such as hurricanes. During hurricane Irma, there were traffic jams from Florida to Tennessee — over 700 miles long. Coordinated routing and information could make evacuation traffic jams a thing of the past.

The broader societal context

I like to think of transportation as a stage for society. Transportation is extremely complex, but simpler than the interconnected complexities of society. Take this example of selfish vs social routing.

In society, the scenarios are not that simple. Capitalism has served us pretty well, even though it is clearly not optimal. Why is this the case? It’s because of a lot of additional complexities, and small perturbations that can have large emergent consequences.

For example, if half of our society believes socially optimal solutions are ‘evil’ because of misinformation, no matter how large the potential for improvement is, it will end up failing. Another issue is the person or group that decides what is socially optimal could have flawed judgement.

But how could big data and information revolutionize our society? During this presidential election cycle, more than 10 billion$ were spent — that’s larger than the GDP of Haiti!

But isn’t the reason we have elections for the people? Do the people benefit from blowing money on TV, Google, and Facebook ads? Rather than spending in this way, big data and scientific methods could make it objectively possible to infer which policies work for the larger societal good through detailed information gathering.

More money can be allocated towards optimal policies rather than infighting between political parties. The catch is that this is a powerful tool, that can also be misused. Transparency, accountability, and big data — the confluences of these 3 could change society for the better.