Uma proposta de fiscalização do ISSQN em Fortaleza

Abstract

The objective of this research is to investigate the behavior of taxpayers of the Imposto Sobre Serviços de Qualquer Natureza (ISSQN) [Tax on Services of any Nature] in relation to their billing, the calculation basis offered for taxation and ISSQN declared monthly to the municipality of Fortaleza. This analysis will indicate unusual changes in the mentioned amounts in the time series corresponding to each taxpayer and will serve as a criterion for selecting, confirming and combating possible cases of tax evasion. For this, the mentioned values were collected, monthly from January 2013 to September 2019, through Secretaria de Finanças de Fortaleza (SEFIN) [Finance Department of Fortaleza], and three taxpayers were randomly chosen from the 100 most expressive, for each of the five most representative segments among the 58 segments of activities created by SEFIN. The methodology used was presented by Chen & Liu (1993), in the article “Joint Estimation of Model Parameters and Outlier Effects in Time Series”, which is used in an automatic procedure for detecting outliers in time series, implemented by López-de -Lacalle (2019), in the tsoutliers package, which can be executed together with the automatic selection strategy of the ARIMA model. The results showed four taxpayers with outliers within the time series. The taxpayers that presented anomalies were those of registration 3, 4, 8 and 12, in which the first is from the engineering segment, the next two are from hosting segment and the last is from technical or administrative support services segment (with significant billing). The inscriptions for the other two segments, hospital and selection and supply of labor, did not present outliers in the observed variables.