Lei de Newcomb-Benford: uma aplicação na conformidade dos registros dos gastos públicos dos municípios tocantinenses

Abstract

The present study aimed to verify empirically whether the application of the NewcombBenford law contributes to the detection of relevant distortions on public expenditures, based on those made by the municipalities of Tocantins, Brazil. Hence, the accounting model developed by Santos et al. (2005), which is based on the relationship between the Newcomb-Benford Law (NB-Law) and the Hypothesis Tests (Z-Test and ChiSquare-test) with the purpose of analyzing the compliance of public expenditures to NB-Law. It also compares the differences in expected probability (Ep) distributions by NB-Law with observed probability (Op) in the analysis of Tocantins cities notes of commitment. The results showed that two of them, which had greater disagreements with the NB-Law, are classifying expenses with labor force as a continuous nature on the element of expense "36 – general outsourcing services – private individual". Such classification is made in order to reducing the index of staff expenses, fitting within the guideline limit stipulated by the LRF. The municipalities complying with the NB-Law have also made mistakes / manipulations that were not evidenced for two reasons: (1) the first position digit of the notes of commitment, which had a mistake/manipulation, were different; (2) mistakes/manipulations were not as widespread as in non-compliant municipalities. The findings of the research showed the usefulness of applying the Newcomb-Benford Law to improve the audit planning process, contributing to the audit track construction as they identify nonconformities of public funds expenditure.