Why urban heterogeneity limits the 15-minute city

This paper demonstrates that the feasibility of the "15-minute city" paradigm is fundamentally constrained by the internal structure of urban economies and firm-size distributions, revealing that universal short commutes are mathematically impossible in highly concentrated employment centers without significant economic restructuring or differentiated mobility strategies.

The Gist

Here is an explanation of the paper "Why urban heterogeneity limits the 15-minute city," translated into simple, everyday language with creative analogies.

The Big Idea: Why You Can't Just "Walk Everywhere"

Imagine the "15-minute city" as a utopian promise: a neighborhood where you can walk or bike to your job, grocery store, doctor, and school in 15 minutes or less. It sounds perfect, right?

This paper argues that while this works great for getting to the bakery or the park, it is mathematically impossible for everyone to walk to their job in many modern cities.

The reason isn't bad planning or slow traffic. The reason is how big companies are structured.

The Core Problem: The "Giant vs. The Minnows"

To understand why, let's look at how jobs are distributed in a city.

The Analogy: The Pizza Party
Imagine a city is a giant pizza party.

• The "15-minute city" dream says: Everyone should be able to walk to a slice of pizza in 15 minutes.
• The Reality: Most of the pizza is eaten by a few huge, hungry giants (large corporations like Amazon, Google, or massive hospitals), while the rest of the guests are small groups of friends (small local businesses) sharing tiny slices.

In the real world, a tiny number of huge companies employ a massive chunk of the workforce. Meanwhile, there are thousands of tiny shops that employ just a few people.

The Mathematical "Traffic Jam"

The author, Marc Barthelemy, uses math to show that this imbalance creates an unavoidable traffic jam, even if we rearrange the city perfectly.

The "Giant" Problem:
Let's say the biggest company in your city employs 10,000 people.

• If this company is located in the center of the city, it needs to pull workers from a huge circle around it to find 10,000 people.
• Even if you place this giant company in the absolute best spot possible, the people living on the edge of that huge circle will have to travel a long way to get there.
• The Catch: Because the company is so big, its "catchment area" (the area it needs to draw workers from) is physically too large to fit inside a 15-minute walking radius.

The "Small" Advantage:
Now, imagine a small coffee shop that employs 5 people.

• It can sit in a tiny corner of the neighborhood.
• It only needs to pull 5 people from a very small area.
• Everyone can walk there easily.

The Conclusion:
If a city has a mix of tiny shops and a few massive employers, the massive employers dictate the rules. No matter how you shuffle the buildings around, those 10,000 workers at the giant company must travel further than 15 minutes because there simply aren't enough people living close enough to fill their seats without stretching the distance.

The "Zipf" Law (The Rule of the Big Fish)

The paper mentions something called a Zipf distribution. Think of this as the "Law of the Big Fish."

• In almost every city, the biggest company is huge, the second biggest is half that size, the third is a third, and so on.
• This isn't random; it's how economies naturally grow.
• The paper shows that when this "Big Fish" gets too big (which happens in big cities like Paris), the math says you cannot fix the commute times just by moving buildings.

The Paris Test Case

The authors tested this theory on Paris, France.

• Paris is a dense, walkable city.
• They asked: "Can we rearrange all the jobs in Paris and its immediate suburbs so that everyone can walk to work in 15 minutes?"
• The Answer: No.
• Even if you put the biggest companies in the exact center and the smallest ones on the edges, the sheer size of the biggest employers means some people have to live far away to work there.
• To make a 15-minute commute possible for everyone, Paris wouldn't just need better bikes or faster trains; it would need to break up its biggest companies or change its entire economic structure so that jobs are spread out more evenly.

What Does This Mean for Us?

This paper doesn't say the 15-minute city is a bad idea. It just says we need to be realistic about what it can achieve.

1. For Daily Life (Groceries, Parks, Schools): The 15-minute city works great! These services are usually small and can be scattered everywhere.
2. For Work (Commuting): The 15-minute city hits a hard wall. Because big companies are so big, they act like a "black hole" that pulls workers from far away.

The Final Takeaway:
We shouldn't ask, "Can we make a perfect 15-minute city?"
Instead, we should ask, "What is the shortest commute time that is actually possible given the size of our biggest companies?"

For some cities, that might be 15 minutes. For big, complex cities like Paris or New York, the "minimum possible" might be 30 or 45 minutes, and that's okay—it's just the cost of having a massive, diverse economy. The solution isn't just walking; it's building better trains and buses to bridge the gap that the "Big Fish" creates.

Technical Summary

Here is a detailed technical summary of the paper "Why urban heterogeneity limits the 15-minute city" by Marc Barthelemy.

1. Problem Statement

The "15-minute city" concept posits that all essential urban functions (work, commerce, healthcare, etc.) should be accessible within a 15-minute walk or cycle. While this paradigm has gained traction for reducing carbon emissions and improving quality of life, its feasibility regarding commuting to work remains largely unexamined quantitatively.

Existing literature often focuses on local amenities (shops, schools) which are easily redistributed. However, employment is structurally different: it is governed by firm-size distributions that are inherently heterogeneous and not freely replicable. The paper addresses a fundamental structural question: Is it theoretically possible to spatially rearrange workplaces in a city such that every worker can reach their job within a prescribed time $\tau_0$, regardless of planning choices or transport technology?

2. Methodology

The author employs a minimal spatial framework combining urban geometry with empirically observed economic statistics to derive strict theoretical lower bounds on commuting times.

A. Theoretical Model

• Urban Geometry: An idealized city is modeled as a disk of area $A$ (radius $R$) with a uniform population density.
• Firm Distribution: Employment is provided by $N$ firms. Firm sizes ($m_r$) follow a Zipf distribution (power law):
  $$m_r \propto r^{-\gamma}$$
  where $r$ is the rank of the firm (1 being the largest) and $\gamma$ is the exponent controlling economic inequality.
  • $\gamma = 0$: Homogeneous firms.
  • $\gamma > 1$: Highly concentrated employment (few large firms, many small ones).
• Optimization Problem: The goal is to find a spatial arrangement of firms that minimizes the maximum commuting distance ($\ell_{max}$) for the entire population. This is modeled as a weighted Voronoi partition problem where firm sizes act as weights.

B. Analytical Derivation

The paper derives a strict lower bound ($L^*_0$) based on the constraint imposed by the largest firm. Even if the largest firm is placed at the point of maximum population density, it must draw a specific fraction of the workforce from a finite catchment area.
The derived lower bound for the maximum commuting distance is:
$$L^*_0 \geq \sqrt{\frac{A}{\pi H_N(\gamma)}}$$
Where $H_N(\gamma) = \sum_{r=1}^N r^{-\gamma}$ is the generalized harmonic number.
The minimal achievable commuting time $\tau_0$ is then:
$$\tau_0(\gamma) = \frac{1}{v} \sqrt{\frac{A}{\pi H_N(\gamma)}}$$
where $v$ is the travel speed.

C. Numerical Validation

• Simulation: The author uses Simulated Annealing to numerically optimize firm locations for finite systems ($E=1000$ to $5000$ employees).
• Parameters: The model tests various $\gamma$ values, total employment $E$, and firm counts $N$ (scaling linearly with $E$).
• Capacity Tolerance: A tolerance parameter $\delta$ is introduced to allow firms to absorb slight mismatches between capacity and assigned workers, simulating real-world flexibility.

D. Case Study: Paris

The model is applied to Paris and its near suburbs (petite couronne), using INSEE data (2023) to estimate the empirical Zipf exponent $\gamma \approx 1.38$. The analysis checks if the 15-minute threshold is achievable for 1.314 million workers in this region.

3. Key Contributions

1. Identification of a Structural Limit: The paper establishes that the feasibility of an $x$-minute city is not merely a function of urban design or transport speed, but is fundamentally constrained by the heterogeneity of firm sizes (economic structure).
2. Theoretical Lower Bound: It provides a closed-form analytical bound showing that for cities with high economic concentration ($\gamma > 1$), the minimum commuting time is independent of the number of firms and depends only on the city's area and the degree of inequality.
3. Phase Transition: The study identifies a sharp transition in feasibility. When $\gamma < 1$, increasing the number of firms can reduce commuting times arbitrarily. When $\gamma > 1$, the system hits a "hard floor" where no spatial rearrangement can achieve universal short commutes.
4. Differentiation of Amenities vs. Employment: It clarifies that while local services can be redistributed to meet proximity goals, employment cannot be decoupled from its underlying economic distribution without restructuring the economy itself.

4. Results

• The Regime of $\gamma > 1$: For empirically observed values of $\gamma$ (typically $0.7$to$1.4$in major metros), the harmonic number$H_N(\gamma)$converges to the Riemann zeta function$\zeta(\gamma)$. Consequently, the lower bound on commuting distance becomes **independent of the number of firms ($N$)**. Adding more small firms does not help the workers of the few massive firms.
• The Paris Stress Test:
  • Using $\gamma \approx 1.38$ for Paris, the theoretical minimum commuting time for walking ($v=5$ km/h) is significantly higher than 15 minutes.
  • Even with cycling ($v=15$ km/h) and generous capacity flexibility ($\delta=30\%$), the 15-minute threshold cannot be reached for the entire workforce.
  • The analysis shows that achieving a 15-minute commute for all workers in Paris would require a massive restructuring of the economic landscape (reducing $\gamma$), not just better planning.
• Numerical vs. Theoretical: Numerical simulations confirm the theoretical bounds. In finite systems, the achievable $\ell_{max}$ is slightly higher than the theoretical asymptotic bound due to discrete effects, but the qualitative "infeasibility" for $\gamma > 1$ holds true.

5. Significance and Implications

• Redefining the Goal: The paper argues that the question should shift from "Can we build a 15-minute city?" to "What is the minimal feasible $x$ given a city's economic structure?"
• Policy Implications:
  • Differentiated Strategy: Proximity-based planning is effective for daily amenities (shops, schools) but insufficient for employment in heterogeneous economies.
  • Mobility vs. Land Use: For large, economically concentrated cities, achieving short commutes for all workers requires complementary mobility solutions (efficient public transport) rather than relying solely on spatial redistribution or the 15-minute walking radius.
  • Economic Restructuring: Truly universal proximity to work in large metropolises would necessitate a fundamental change in the firm-size distribution (e.g., breaking up large monopolies or decentralizing large employers), which is a political and economic challenge beyond urban planning.

In conclusion, the paper demonstrates that urban heterogeneity acts as a fundamental physical constraint on the 15-minute city paradigm. For cities with high economic concentration, the 15-minute ideal is structurally unachievable for commuting without altering the underlying economic organization of the city.