Empirical Economics Социальные науки (прочее)
Equivalent contamination proportion (ECP): interpreting deviations from Benford’s law in economic data beyond statistical significance
Аннотация
Conformity with Benford’s law is widely used in accounting and finance to identify data irregularities, detect potential fraud, and evaluate the reliability of reported numerical information. However, the statistical tests commonly used for this purpose suffer from three well-known limitations: excessive sensitivity to sample size, limited interpretability, and the absence of a unified metric that enables consistent comparisons across tests and empirical contexts. This paper introduces the equivalent contamination proportion (ECP), an interpretative, benchmark-calibrated measure designed to complement, rather than replace, traditional Benford conformity tests. The ECP is defined as the proportion of benchmark contamination in a hypothetical Benford-conforming sample that would generate the same expected divergence as that observed in the empirical data. It is a continuous and easily interpretable measure ranging from 0 to 100%. A key feature of the ECP is that its interpretation is not driven by dataset size, which allows more meaningful comparisons of the magnitude of deviations across datasets. In addition, it yields coherent results across commonly used divergence statistics, thereby facilitating comparisons across metrics and empirical settings in financial research. Closed-form and simulation-based methods for estimating the ECP from the most widely used statistics are described in this paper. A retrospective analysis of three influential studies illustrates how the ECP enhances the interpretability of Benford conformity results in accounting and financial research.