ABSTRACT

This study uses a decade of financial accounting data to examine if and how they depart from Benford's Law. Using a large sample of U.S. public companies, we conduct an analysis of the first-two digits of data items generally used in research to measure total accruals and discretionary accruals and where fraud, restatements, and enforcement actions are revealed. We break down a decade of data into six subperiods; pre-SOX Period (2001), SOX 1 Period (2002–2003), SOX 2 Period (2004–2006), SOX 3 Period (2007), Crisis 1 Period (2008), and Crisis 2 Period (2009–2010). We find different indicators of manipulation during the periods studied, as well as differences between small and big companies and companies audited by Big 4 and non-Big 4 firms.

You do not currently have access to this content.