The Braess Paradox: Why More Roads Mean More Traffic Jams

A professional 15-slide deck exploring the counterintuitive Braess Paradox — why adding roads can worsen congestion — covering its mathematical origins, real-world case studies, and implications for urban planning and network design.

More Roads, More Traffic

The Braess Paradox and Why Everything We Know About Congestion Is Wrong

Roadmap

The Counterintuitive Truth

How the Paradox Works

Origins: From Pigou to Braess

Real-World Evidence

Beyond Roads

Implications for Tomorrow's Cities

02

THE PARADOX

Sometimes the best way to fix traffic is to remove a road.

Adding a Road Can Make Everyone Slower

Braess's Paradox: increasing network capacity can decrease overall performance for every single user

First observed by economist Arthur Pigou in 1920; formalized by mathematician Dietrich Braess in 1968

Occurs when independent drivers optimize their own routes selfishly — a Nash equilibrium emerges

The system stabilizes at a point where no individual can improve — but everyone would benefit from cooperation

The paradox flips intuition: removing roads can be the optimal strategy to improve total flow

Sources: 1, 2

The Classic 4-Node Network

Picture two routes from Start to End, each with a fixed-cost segment and a congestion-sensitive segment

Adding a zero-cost shortcut between the two midpoints creates an irresistible third path

Every driver rationally switches to the shortcut — overloading it until all three routes are slower than before

The shortcut destroys the natural load-balancing that existed in the original two-route system

Result: everyone's travel time increases, even though the network gained a new road

Nash Equilibrium vs. Social Optimum

Nash Equilibrium

Social Optimum

Each driver independently selects the fastest-looking route. No single driver can reduce their travel time by switching alone. The system freezes at a stable point — but it is not efficient. Collectively, everyone loses time compared to a coordinated outcome.

A central planner assigns routes to minimize total system travel time. Some individuals may travel slightly longer than they would prefer — but the aggregate outcome is faster for the entire population. The gap between these two states is the cost of selfish routing.

06

ORIGINS

A mathematical curiosity that took 40 years to reach the mainstream.

From Theory to Reality

1920

1968

1990s

2005

2010s

Arthur Pigou identifies congestion externalities

Dietrich Braess publishes the paradox (in German)

Computational models confirm the paradox at scale

English translation appears in Transportation Science

Real-world validations: Seoul, San Francisco, New York

Sources: 1, 7, 8

Dietrich Braess and His 1968 Paper

In 1968, Dietrich Braess, a mathematician at Ruhr University in Germany, published 'Über ein Paradoxon aus der Verkehrsplanung' — a concise paper demonstrating that adding a road to a network could increase every driver's travel time. The work circulated only among German-speaking mathematicians for decades. It wasn't until November 2005, when the journal Transportation Science published an English translation, that the paradox exploded into global awareness among urban planners, computer scientists, and economists. Braess himself has noted that he was surprised by how long it took for the practical implications to be recognized. Today, his paper is one of the most cited works in transportation science and network theory, referenced across fields ranging from internet routing to metabolic biology.

Sources: 7, 8

Let's spend a moment on the man and the paper. Dietrich Braess was not a transportation engineer — he was a pure mathematician working on numerical analysis. His 1968 paper was remarkably short, barely a few pages, and written entirely in German. For context: it was the height of highway-building optimism, when cities around the world were carving expressways through neighborhoods, convinced that more asphalt meant faster commutes. Braess's counterintuitive proof sat quietly, almost ignored, for four decades. When the English translation finally landed in 2005, it landed in a very different world — one dealing with chronic congestion, armed with GPS data and simulation tools. The paper's timing, ironically, was perfect: it arrived exactly when cities were starting to question whether the highway era had been a mistake.

09

REAL-WORLD EVIDENCE

When cities closed roads, traffic didn't collapse — it vanished.

Seoul: Cheonggyecheon Restoration

An elevated highway was demolished. A buried river was restored. Traffic improved.

Sources: 9

In 2003, Seoul did something radical. They demolished the Cheonggyecheon elevated highway — a six-lane expressway carrying 168,000 vehicles daily through the heart of the city — and restored the buried river beneath it. Critics predicted paralyzing gridlock. Instead, the opposite happened. Traffic in the surrounding area actually improved. Drivers adapted: they changed routes, shifted travel times, and many switched to the expanded public transit system. Meanwhile, the restored river became one of Seoul's most beloved public spaces, attracting millions of visitors, spurring business growth, reducing the urban heat island effect by up to nearly 6 degrees Celsius, and increasing biodiversity. It is the Braess Paradox made visible — a real-world demonstration that removing capacity can heal both traffic and the city.

5.9°C

Maximum reduction in urban heat island effect after Cheonggyecheon highway removal, Seoul

Sources: 9

Beyond Roads: The Paradox Everywhere

Electrical power grids: adding transmission lines can destabilize the network and trigger cascading blackouts

Biological systems: extra metabolic pathways can reduce overall efficiency in cellular networks

Internet routing: more peer connections between autonomous systems can increase latency for all users

Sports analytics: removing a star player can sometimes improve team performance — the 'Ewing Theory'

The paradox emerges whenever decentralized agents compete for shared, congestible resources

Sources: 1, 3, 8

What This Means for Urban Planners

Building more lanes is not a long-term congestion solution — induced demand compounds the Braess effect

Removing redundant or underused road links can unlock better system-wide performance

Invest in public transit, cycling infrastructure, and pedestrian-friendly design instead of road expansion

Congestion pricing can align individual incentives with the social optimum — London and Stockholm prove it

Simulate network changes rigorously before breaking ground; computational modeling is essential

Key Takeaways

The Braess Paradox proves that adding road capacity can slow everyone down when drivers route selfishly

The phenomenon is not theoretical — it has been confirmed in simulations and real-world urban case studies

Seoul's Cheonggyecheon and San Francisco's Embarcadero Freeway removal both improved traffic flow

The paradox extends far beyond roads: power grids, internet routing, and biological networks all exhibit it

Smarter cities design for the system optimum — investing in coordination, transit, and network analysis

References

[1] Braess's paradox - Wikipedia — en.wikipedia.org [2] Braess's paradox - Wikipedia — en.wikipedia.org [3] Exploring the Braess Paradox: Static Versus Dynamic Assignment | Collective Dynamics — collective-dynamics.eu [7] [PDF] Debunking Braess' Paradox - ITE Western District — westernite.org [8] The Braess Paradox — supernet.isenberg.umass.edu [9] Roads, traffic and Braess’s paradox – rachel.blog — rachel.blog Braess' paradox — politesi.polimi.it [PDF] Braess's Paradox in Large Random Graphs - Stanford CS Theory — theory.stanford.edu Informational Braess' Paradox — economics.mit.edu

More AI presentation examples · Create your own with Sliderest