Skip to main content
Category: Paradoxes
Type: Mathematical/Network Paradox
Origin: 1968, Dietrich Braess (German mathematician)
Also known as: Braess Paradox, Network Paradox
Quick Answer — Braess’s Paradox demonstrates that adding a new path to a network can make every individual worse off, even when no one acts selfishly. First identified by German mathematician Dietrich Braess in 1968, this paradox shows that optimized systems can be counterintuitively fragile—adding capacity sometimes reduces overall efficiency.

​
What is Braess’s Paradox?

Braess’s Paradox is one of the most counterintuitive discoveries in network theory and transportation planning. It shows that in a system where individuals act rationally to minimize their own travel time, adding a new route can paradoxically increase travel time for everyone—even though the new route seems to offer a shortcut.
“In traffic networks, the whole is often less than the sum of its parts.” — Dietrich Braess
The paradox operates on a fundamental principle of game theory: in a Nash equilibrium, no individual can improve their situation by unilaterally changing their strategy. However, when a new option is added, the equilibrium shifts, and the new “optimal” choice for each individual may lead to worse collective outcomes.

​
Braess’s Paradox in 3 Depths

  • Beginner: Imagine two routes from A to B, each taking 45 minutes with moderate traffic. A shortcut opens that seems to take 15 minutes—but when everyone takes it, both routes become jammed and everyone takes longer than before.
  • Practitioner: This explains why adding lanes to highways sometimes worsens congestion, why new subway lines can make overall commute times longer, and why urban planners must carefully model network effects before building new infrastructure.
  • Advanced: The paradox reveals deeper truths about emergent systems: local optimization doesn’t guarantee global efficiency. It’s a special case of the price of anarchy—the inefficiency that arises when individuals don’t coordinate their actions.

​
Origin

Braess’s Paradox was discovered by German mathematician Dietrich Braess in 1968. While working at the Ruhr University Bochum, Braess was studying traffic flow and network optimization when he noticed this counterintuitive phenomenon. His original example used a simple network: two roads from Start to End, each with two segments. Drivers could choose either route. Then Braess added a “shortcut” road that connected the middle of one route to the middle of another. The mathematical analysis showed that this new road—which seemed like it should help—would actually increase travel time for all drivers. The paradox was largely forgotten until the 2000s, when real-world examples began appearing. Cities that added roads to reduce congestion sometimes found traffic getting worse. The phenomenon has since been studied in power grids, biological networks, and even the mechanics of human movement.

​
Key Points

1

Selfish Routing Drives the Paradox

Each driver chooses the fastest route for themselves. When a new road appears, drivers recalculate—but the new “best” choice for individuals creates worse outcomes for everyone.
2

The Network Must Be Unbalanced Initially

The paradox only occurs when the network is not already at optimal flow. Well-designed networks where routes are already balanced typically won’t experience this phenomenon.
3

Removing Roads Can Improve Flow

In some cases, closing a road can actually improve overall traffic flow—the reverse of what intuition suggests. This has been observed in real cities.
4

The Price of Anarchy

Braess’s Paradox is a specific case of the broader “price of anarchy”—the ratio between optimal coordinated outcomes and selfish individual choices in non-cooperative games.

​
Applications

Urban Transportation Planning

Cities must model network effects before adding roads. Simply building more infrastructure doesn’t guarantee improved flow—careful analysis of how drivers will respond is essential.

Internet Routing

The paradox applies to data networks too. Adding faster links or more bandwidth can sometimes worsen overall latency if routing algorithms aren’t properly coordinated.

Power Grid Management

Electrical grids can exhibit similar paradoxes. Adding transmission lines to balance load can sometimes create new bottlenecks elsewhere in the system.

Economic Policy

Markets sometimes exhibit analogous behavior—adding options can reduce overall welfare when individuals make suboptimal choices in isolation.

​
Case Study

The most famous real-world demonstration of Braess’s Paradox occurred in Stuttgart, Germany, in 1969. City planners had added a new road to relieve traffic congestion in the city center. Instead, traffic jams became worse. Years later, in 1990, the phenomenon was observed again when a section of the 42nd Street in New York City was closed for renovations. Traffic flow actually improved—closing a road made commutes faster. Similar observations have been made in Seoul, South Korea, where the removal of a major highway (the Cheonggyecheon restoration) improved overall traffic despite reducing road capacity. These cases confirm Braess’s mathematical insight: networks can be counterintuitive, and adding capacity doesn’t always mean better performance.

​
Boundaries and Failure Modes

Braess’s Paradox has several important limitations:
  1. Requires selfish optimization: The paradox assumes drivers (or network users) act purely in self-interest. If everyone perfectly coordinates, the paradox disappears.
  2. Only applies to certain network topologies: Not all networks exhibit the paradox. It requires specific configurations where the new path creates a “bottleneck” that didn’t exist before.
  3. Assumes fixed demand: The paradox assumes the number of users stays constant. If adding a road attracts more users (induced demand), the analysis becomes more complex.
  4. Time-horizon matters: In the short term, the paradox may hold. Over longer periods, users may change behavior patterns, relocate, or find alternative solutions.

​
Common Misconceptions

Braess’s Paradox shows that more roads can mean worse traffic. The relationship between capacity and performance is non-linear and depends on network structure.
The paradox is mathematical and applies to any network where users choose routes selfishly—data networks, power grids, and even biological systems can exhibit this behavior.
The paradox doesn’t mean infrastructure is useless—it means careful modeling is essential. Most networks don’t exhibit the specific conditions needed for the paradox.

Nash Equilibrium

A situation where no player can improve their outcome by unilaterally changing strategy—the foundation of Braess’s Paradox.

Price of Anarchy

The ratio between optimal coordinated outcomes and selfish individual choices in distributed systems.

Induced Demand

The phenomenon where building more roads actually increases total traffic volume by attracting new users.

​
One-Line Takeaway

Braess’s Paradox teaches us that in complex systems, adding capacity doesn’t guarantee improvement—local optimization can produce global inefficiency, and careful network analysis is essential before building new infrastructure.