Benford’s Law Explained: Why Many Numbers Start With 1

Quick answer: Benford’s law says that in many real-world data sets, the first digit is more often small: about 30.1% of numbers start with 1, 17.6% with 2, and only 4.6% with 9. The probability that the first digit is d is log₁₀(1 + 1/d). It holds for data that span several orders of magnitude, such as populations, river lengths and financial records, and auditors use deviations from it to flag possible fraud. Simon Newcomb noticed it in 1881 and Frank Benford tested it in 1938.

If you pick a random number from a newspaper, a tax return or a list of city populations, what is the chance it starts with 1? Intuition says one in nine. In many such data sets it is about 30%. This surprising pattern is called Benford’s law.

The Law

The probability that the first (leading) digit is \(d\) is

\[ P(d) = \log_{10}\left(1 + \frac{1}{d}\right). \]

First digitProbability
130.1%
217.6%
312.5%
49.7%
57.9%
66.7%
75.8%
85.1%
94.6%

Need a refresher? See our guide to logarithms.

A Quick Example: Powers of 2

Look at \(2, 4, 8, 16, 32, 64, 128, 256, 512, 1024, \dots\). Count the leading digits of the first 1,000 powers of 2:

First digit123456789
Count (of 1,000)301176125977969565245

That matches the Benford table almost exactly. Small samples are noisier: among only the first 50 powers, the counts are 15, 10, 5, 5, 5, 4, 1, 5 and 0. Why do 1s win? Doubling from 100 takes you to 200, so a number spends a whole doubling step among the “1s”. Getting from 900 to 1000 only needs a factor of \(\frac{10}{9}\), about 11% growth, so numbers pass through the “9s” quickly. Not every growing sequence works, though: powers of 10 always start with 1.

Why It Happens

Benford’s law appears when data are spread over several orders of magnitude and are roughly evenly spread on a logarithmic scale. On a log scale, the stretch from 1 to 2 takes up \(\log_{10} 2 \approx 30.1\%\) of each decade, while 9 to 10 takes only 4.6%. The Benford distribution is also scale invariant: if every value in a Benford data set is multiplied by the same positive constant, such as converting dollars to euros or meters to feet, the overall distribution stays the same, even though individual leading digits change. It is the only first-digit law with this property, a strong hint that it is natural.

Where It Works, and Where It Does Not

Often close to Benford (check first)Usually does not fit
Populations of many towns, river drainage areas, many accounting amounts, many physical constantsAdult heights, test scores, phone numbers, assigned IDs, prices set at 9.99, data with minimums, maximums or caps

A quick check before you rely on it: the data should span several orders of magnitude, come from natural processes rather than assigned numbers, and have no built-in limits. Even then, a particular data set may not fit.

Catching Fraud

People who invent numbers often, but not always, spread first digits too evenly; someone who knows about Benford’s law can also fake data that fits it. Forensic accountants such as Mark Nigrini popularized Benford tests to flag suspicious accounts and tax returns, and it has been accepted as supporting evidence in some court cases. A mismatch is a red flag, not proof: many honest data sets do not follow the law. Claims that Benford’s law can detect election fraud are especially controversial, because precinct vote counts often do not fit it naturally.

History

Astronomer Simon Newcomb noticed in 1881 that the early pages of logarithm tables, used for numbers starting with 1, were more worn than later pages. Physicist Frank Benford rediscovered it in 1938 and tested it on over 20,000 numbers from 20 data sets, so the law bears his name.

Common Misconceptions

“All data follow Benford’s law.” No. It needs data spanning many orders of magnitude without artificial limits.

“A Benford mismatch proves fraud.” No. It only suggests where to look more closely.

Watch: Benford’s Law in Video

If you learn better by watching, these videos are worth your time.

Number 1 and Benford’s Law – Numberphile (Numberphile). A clear explanation of why 1 appears so often.
Benford’s Law in Real Life (Minding The Data). Testing the law on real-world data sets.

Check Your Understanding

Five quick questions to see how much of this has stuck. Pick an answer and press Submit; you will see a short explanation either way.

Answer each question and press Submit to see your result right away.

1. About what percentage of numbers start with 1 under Benford’s law?
Found a bug? Let us know!
2. What is the probability of a leading 9?
Found a bug? Let us know!
3. Which data set is least likely to follow Benford’s law?
Found a bug? Let us know!
4. Who first noticed the pattern in 1881?
Found a bug? Let us know!
5. Does a mismatch with Benford’s law prove fraud?
Found a bug? Let us know!

Frequently Asked Questions

What is Benford’s law?

An observation that in many real-world data sets, leading digits are more often small: about 30% start with 1 and under 5% start with 9.

What is the formula for Benford’s law?

The probability that the first digit is d equals log₁₀(1 + 1/d), for d = 1 to 9.

Why does Benford’s law work?

It holds when data span several orders of magnitude and are roughly uniform on a logarithmic scale, where the interval from 1 to 2 covers about 30% of each decade.

How is Benford’s law used to detect fraud?

Auditors compare the first digits of financial data with Benford’s distribution. Large deviations flag records for closer review, though they do not prove fraud, and careful fakers can produce data that fits the law.

Does every data set follow Benford’s law?

No. It fits best when data span several orders of magnitude without built-in limits. Heights, test scores, phone numbers and capped or assigned values usually do not follow it.

Keep Exploring

Sources and Further Reading

Related to This Article

What people say about "Benford’s Law Explained: Why Many Numbers Start With 1 - Effortless Math"?

No one replied yet.

Leave a Reply