ICMS forecast in the states of the Southeast region: use of the simple seasonal model and Holt-Winter

Abstract

As a way to improve the problems arising from the Bureaucratic Management Model, the Management Model of Public Administration incorporated the Principle of Efficiency into the state machine and, with this, private initiative practices began to be used by the public administration. One of these practices was planning, which is pointed out as a starting point to obtain efficiency (SILVA et al., 2013) and which consists of anticipating the future in order to make the best decisions in the present. In Public Administration, planning and forecasting of revenue gain importance, as it is from this that expenditures in the public sector are fixed. Therefore, this article aimed to identify, based on the Time Series models, which is the best for forecasting ICMS tax revenue for the States of the Southeast Region (MG, SP, RJ and ES). The time period corresponding to the period from 2010 to 2017 was used for the projections, and the IBM SPSS software, version for students, was used for their analysis. The Additive Holt-Winters models for the States of MG, SP and RJ and the Simple Seasonal Model for the State of Espírito Santo were appointed. In this way, it was identified that the Time Series can present a resource to assist the Public Administration in forecasting ICMS revenues for the States of the Southeast Region.