© D.L. Crumbley
When Benford Analysis Is or Is Not Likely Useful
Thefts, kickbacks, contract rigging.
Where no transaction is recorded.
Set of assets that must meet a threshold to be recorded.
Accounts with a built in minimum or maximum.
An account specifically set up to record $100 refunds.
Accounts with a large number of firm-specific numbers.
Prices set at psychological thresholds ($1.99), ATM withdrawals.
Numbers that are influenced by human thought.
Check numbers, invoice numbers, zip codes.
Data set is comprised of assigned numbers
When Benford Analysis Is Not Likely Useful
Most sets of accounting numbers.
Accounts that appear to conform – When the mean of a set of numbers is greater than the median and the skewness is positive.
Full year’s transactions.
On large data sets – The more observations, the better.
Disbursements, sales, expenses.
Transaction-level data – No need to sample.
Accounts receivable (number sold times price). Accounts payable (number bought times price).
Sets of numbers that result from mathematical combination of numbers-Result comes from two distributions.
When Benford Analysis is Likely Useful
Source: Durtschi, Hillison, and Pacini, p. 24.