Benford's Law Applied to Financial Statements

Benford's Law says that in many collections of real-world numbers — city populations, river lengths, or line items on a balance sheet — the first digit is 1 about 30% of the time, 2 about 18%, and so on down to 9 about 4.6%, rather than each digit appearing equally often. This isn't a coincidence about the number 1; it falls out of how numbers that span many orders of magnitude are distributed.

Forensic accountants apply this by tabulating the leading digit of thousands of numbers from a company's ledger — invoice amounts, expense entries, journal postings — and comparing the observed frequencies to Benford's expected curve using a chi-squared or similar statistical test. Numbers that are fabricated by a human, rather than arising from a genuine business process, tend to be distributed too evenly across digits, because people are bad at mimicking Benford's skew when making numbers up.

Worked example. A regulator examines 5,000 expense reimbursement entries and finds that only 12% start with 1, versus the ~30% Benford's Law predicts, while entries starting with 4 and 5 are overrepresented — consistent with someone padding numbers to fall just under approval thresholds like $500. That deviation triggers a targeted audit of the flagged entries rather than an automatic finding of fraud.

The test works best on large, naturally generated datasets and is far less reliable on small samples or numbers that are constrained by design, such as prices set at round figures like $9.99 or $19.99.