Contas Anômalas: A Lei de Newcomb-Benford na Análise de Riscos de Irregularidades em Prestações de Contas Eleitorais

Abstract

The main aim of this research was to investigate the possibility of identifying signs of irregularities or deviations in the provision of electoral accounts through an analysis of the distribution of the first digits, based on the Newcomb-Benford Law, and its correlation with the outcomes of the judgment processes. The theoretical frameworks employed to interpret the context included Agency Theory and concepts of Accountability and Social Control. To achieve the proposed objectives, 661,144 electoral expenses paid and declared to the Electoral Justice by 5,532 candidates for federal deputy in the 2018 Brazilian elections were analyzed in a quantitative analysis stage. This involved comparing the distributions of the first digits of the reported values with the expected proportions according to the Newcomb-Benford Law, using statistical tests (Z, Chi-Square, and Mean Absolute Deviation). In a second stage, a questionnaire was administered to professionals experienced in the analysis and judgment processes of electoral accountability to capture their perceptions of the system, its weaknesses, and openness to innovations. The 187 responses received underwent content analysis. The quantitative results revealed that, in general, electoral accountability does not conform to the expected standards of the Newcomb-Benford Law, irrespective of approval or disapproval by the Electoral Justice, with labor expenses showing particular emphasis, starting with the digit 5 and repeated values of R$500, thereby raising suspicions of possible irregularities undetected by the current system. The questionnaire responses complemented and supported the preceding findings, with deviations related to labor hiring being considered the most challenging to detect by participants. Additionally, the system was described as primarily focused on document verification, with limited inspection initiatives and audits in electoral committees, while civil servants were perceived to have limited material resources and time to perform their duties. Furthermore, the Electoral Justice was recognized as open to innovations deemed beneficial. Consequently, the proposed thesis was partially accepted. These results contribute to a deeper understanding of the adjudication process of electoral accounts in Brazil, while also exploring the application of the Newcomb-Benford Law in an underexplored context, alongside qualitative methods.