Mass Evaluation of Rural Buildings Through Classical Regression and Geostatistics

Abstract

The concept of Rural Territorial Tax (ITR) is rooted in subjective methods that turn this tribute inefficient and inconsistent with their original purposes. Thus, this work had the objective to evaluate the hypotheses that the mass evaluation of rural land can be efficiently used to obtain the ITR, aiming at a more transparent process of taxation with a plausible taxable base. Evaluation methodologies preconized by the Brazilian Normalization NBR 14.653:3 (ABNT, 2004) were applied to data and maps using multiple regression along with geostatistics to generate the standard ground value of rural land. The study area was the Sugarcane Irrigation and Drainage Project-PROJIR, located at the Northern of Rio de Janeiro State, Brazil. The metrics recommended by the International Association of Assessing Officers-IAAO (2013) were used to analyze the performance of land price evaluation. The results indicated that for the multiple regression, the median of the evaluation ratios (0.98) was within the established interval (0.90 a 1.10), while the coefficient of dispersion-COD was 24.60% and the Price Related Differential-PRD was 1.23, which were higher than the recommended values (20% and 1.03, respectively). This indicates that, for the PROJIR region, would be necessary a higher quantity of samples to reach the intervals stipulated by the standards. According to the results obtained in the statics tests of multiple regression, and the space dependency by the semivariograms analysis, this concludes the use of multiple regression in conjunction with the ordinary kriging, in the mass evaluation of rural properties, is viable for purposes of taxes and other applications.