A gentle dive into Benford’s Law and the slow-burn brilliance of Newcomb and Benford
Welcome to FreeAstroScience.
If you flip through the populations of cities, the lengths
of rivers, stock prices, or constants in physics, you might expect the digits 1
through 9 to appear as leading digits about equally. But they don’t. In dataset
after dataset, 1 dominates, while 9 barely shows up.This curious, almost unsettling pattern is known as Benford’s
Law, and once you see it, you never quite forget it.
I’m Flávia Ceccato, and in Brazil, I was recognized by Rank
Brasil as the first person to apply Benford’s Law to fraud detection in public
works audits, earning me a Brazilian record. I’m also co-author of the book
“Seleção de Amostra de Auditoria de Obras Públicas pela Lei de Benford” (Sample
Selection for Public Works Auditing Using Benford’s Law), published by the
Brazilian Institute of Public Works (IBRAOP). With this background, I’d like to
show you how a simple numerical law can reveal much more than it seems.
A Worn Logarithm Table and a Quiet Spark of Genius
To understand where the story begins, we travel back to the
19th century, a time when calculations were done not by apps or machines, but
by people flipping through large printed tables of logarithms.
Canadian-American astronomer and mathematician. Buried in his daily work, he
noticed something odd:
to numbers beginning with 1, were more worn than the rest. Pages for
numbers starting with 8 or 9 looked almost new.
Why were people looking up numbers beginning with 1 more
often?
Newcomb published a short note in 1881 suggesting that many
natural datasets don’t distribute their leading digits uniformly. He even
sketched the basic logarithmic idea behind the phenomenon. And then… the world
moved on. No fanfare. No revolution. The idea simply faded.
Benford Picks Up the Thread
Decades later, in the 1930s, Frank Benford, an
American physicist and engineer, stumbled upon the same mystery. But unlike
Newcomb, he attacked it with data, mountains of it.
He gathered approximately 20,000 real-world numbers from a
diverse range of sources, including river lengths, street addresses, scientific
constants, population data, and even magazine issue numbers. And again, the
same pattern emerged. The digit 1 appeared as the first digit far more
frequently than any other.
Benford formalized the pattern with the now-famous formula:
should appear.
name.
Ironically, neither Newcomb nor Benford lived to see their
work become central to modern data science, auditing, and digital forensics.
Their brilliance bloomed late, long after the world had caught up to it.
What the Law Actually Says
Benford’s Law predicts that the first digit of many
naturally occurring numbers follows a logarithmic pattern rather than a uniform
one. That means:
- About
30% of numbers start with 1 - Only
about 4–5% start with 9
This runs completely against our intuition, but it shows up
whenever data span several orders of magnitude.
Where the Law Appears in Real Life
Benford’s Law surfaces in places where numbers grow, scale,
or expand multiplicatively:
- Economics
& finance: company revenues, stock prices, tax declarations - Nature:
river lengths, earthquake magnitudes, radioactive half-lives - Science
& engineering: large sets of physical constants or measurements
It doesn’t work for constrained or human-assigned
numbers, such as phone numbers, birthdays, or heights in centimeters.
Why It Works (the Intuitive Version)
If you think in percentages rather than absolute amounts,
something interesting happens. A number has to “travel” through a longer range
to go from 1 to 2 than it does from 8 to 9. On a logarithmic scale, this
unevenness becomes clear: numbers spend more “time” beginning with 1.
This property, called scale invariance, is part of
what makes Benford’s Law so universal and so resistant to changes in units.
Whether you express something in metres or miles, dollars or euros, the pattern
often remains.
Catching Fraud With Digits
In modern times, Benford’s Law gained fame not in astronomy
or physics, but in accounting and anti-corruption efforts.
Auditors analyse the first digits of thousands of financial
transactions and compare them with Benford’s distribution:
- If
the digits match the pattern → nothing suspicious - If
they deviate strongly → possible manipulation
It’s a red flag, not a verdict. But it has helped uncover
anomalies in taxes, procurement, contracts, and corporate accounts. Humans tend
to invent numbers too “evenly”; reality does not.
Misconceptions and Limitations
Despite its elegance, Benford’s Law isn’t magic. It fails
when:
- data
have a narrow range - numbers
follow human rules (like assigned codes) - minimums
or maximums distort the distribution - rounding
or reporting thresholds interfere
Misuse of the law has led to unfavorable headlines and
misinterpretations. A good analyst always asks how the data were
generated before reaching for Benford’s curve.
A Slow-Burn Revolution
Newcomb’s worn pages and Benford’s stacks of numbers only
gained their whole meaning in the digital age. Once computers arrived, along
with global finance, big datasets, and sophisticated auditing, their insights
began to emerge.
They were thinkers ahead of their time, seeing structure in
places the world wasn’t yet ready to examine.
A Final Reflection
Benford’s Law teaches us that reality hides patterns in
plain sight. That behind the chaos of numbers lies a subtle order, not obvious,
not intuitive, but deeply rooted in how the world grows and changes.
The next time you scroll through a dataset or glance at a
column of numbers, you might find yourself pausing on the leading digit. And in
that moment, you’ll be sharing a quiet thought with two long-gone scientists
who noticed something most people never see.
This article was written for you by FreeAstroScience.com,
where complex ideas are turned into simple, human stories. Stay curious and
keep watching for the hidden patterns that shape our world.
