Power Laws: The Real Shape of Inequality

"The world's most consequential distributions aren't bell curves. They're power laws. And almost no one was taught to recognize one."

The Citizen Knows This Shape Already

You have lived inside power-law distributions your whole life:

Wealth: the top 0.1% holds far more than 0.1% of the wealth.
Book sales: a handful of bestsellers outsell the rest of the catalog combined.
City populations: Tokyo and New York dwarf the median city.
Social media followers: a few accounts have millions; most have under a hundred.
Network traffic: a few packets carry most of the bytes.
Earthquake magnitudes: most are tiny; a few are catastrophic.

Each one is the same animal: a distribution where the probability of a value decays as a power of the value (P(x) ~ x⁻ᵃ for some constant a), rather than as an exponential. Power-law tails decay slowly. Extreme events are not rare.

The 80/20 Headline, Mathematically

The 'Pareto 80/20 rule' you've heard ('80% of the wealth held by 20% of the people') is the citizen-facing label for one specific power law: the Pareto distribution. The rule is approximate; the actual ratio depends on the tail's specific α. But the structure is universal: a thin slice of the population accounts for a thick slice of the total.

The deeper version: in a true power-law distribution, the rule recurses. Among the top 20%, another 80/20 split applies; among the top 4%, another. Self-similar at every scale. This is what scale-free means in statistics — there is no characteristic 'typical' value the way there is for a bell curve.

Mean and Variance Become Slippery

For a power law with α ≤ 2, the variance is theoretically infinite. For α ≤ 1, even the mean is. 'Infinite' in this context means: as you collect more samples, your estimate of the mean keeps growing rather than settling. The citizen-relevant version of this is the move from the previous lesson: the mean of a right-skewed distribution is misleading; the mean of a true power law is sometimes meaningless.

This matters because so many quantities the citizen acts on (wealth, network influence, drug responses, asset returns) are power-law. Reporting their mean is statistical malpractice; reporting their median and their tail percentiles is the honest move.

The Frame to Internalize

A power law is what inequality looks like when you draw it. The bell curve assumes a 'typical' value exists; the power law denies it. When you can name the shape, you stop being surprised that the top fraction owns most of the system — that's not a bug or a moral failing, it's what power-law mathematics produces. The choice is what you do about it once you can see it.

Code

Top-X% share in a Pareto (power-law) distribution·python

import numpy as np
rng = np.random.default_rng(15)

# Pareto-distributed sample with shape parameter alpha=1.5.
N = 100_000
alpha = 1.5
x = (rng.pareto(alpha, size=N) + 1)   # Pareto draws + 1 (xmin = 1)

# Compute the share of total held by the top 1%, top 10%, top 20%, top 50%.
x_sorted = np.sort(x)[::-1]
total = x.sum()
for pct in (0.01, 0.10, 0.20, 0.50):
    k = int(N * pct)
    share = x_sorted[:k].sum() / total
    print(f"Top {pct*100:>4.1f}%  holds  {share*100:>5.1f}% of the total")

# A 1.5-Pareto shows roughly: top 1% holds ~25%, top 20% holds ~75%.
# The exact numbers depend on alpha. The shape is universal: a few hold most.
# A bell curve cannot produce this shape no matter what mean and std you choose.

Exercise

List three quantities in your own life or work where you suspect a power-law shape (social-media engagement, project effort distribution, customer spend, deep-work session length, web traffic on your site). For each, ask: if I knew the top 1% of values, would I learn more than from the mean? If yes, you're looking at a power-law-shaped reality, not a bell-shaped one.

Hint

Power-law thinking: 'where is the tail, and how much of the total is in it?' Bell-curve thinking: 'where is the center, and how wide is the spread?' They are different mental models, and most citizens default to the wrong one.

The Citizen Knows This Shape Already

The 80/20 Headline, Mathematically

Mean and Variance Become Slippery

The Frame to Internalize

Code

External links

Exercise

Progress

Comments 0