Using Brazil’s Survey of Household Budget (POF) 2017–2018, this study develops a logit model to identify the social groups and their property features that are less likely to pay Urban Property Tax (UPT). The data shows that frequency of UPT payments among urban owner-occupied properties has increased from 36.1% to 56.1% over 15 years (comparing POF 2002–2003 and POF 2017–2018). The logit model reveals that property attributes (regional location, state capital location, and the availability of urban services) were the main determinants of UPT payment. Houses located in Northern Regions were 62.9% less likely of UPT payment, while houses locate in state capitals were 48.0% less likely. Houses benefited of only three typical urban services were 45.4% less likely comparing to houses with five urban services. In addition, households where the head of families are woman, non-white, self-employed or not working, with concern of food shortage, with more children and fewer years of study were also less likely of UPT payment than others. The study concludes that marginalization and regional specificities were the main determinants of UPT payment propensity and recommends tax policies to ensure more progressivity and to reduce regional disparities.

本研究使用巴西 2017-2018 年家庭预算调查 (POF)，开发了一个 logit 模型来识别不太可能支付城市财产税 (UPT) 的社会群体及其财产特征。数据显示，15 年来，城市自有物业的 UPT 支付频率从 36.1% 增加到 56.1%（比较 POF 2002-2003 和 POF 2017-2018）。logit 模型显示，房产属性（区域位置、州首府位置和城市服务的可用性）是 UPT 支付的主要决定因素。位于北部地区的房屋支付 UPT 的可能性降低了 62.9%，而位于州首府的房屋则降低了 48.0%。与具有五种城市服务的房屋相比，仅受益于三种典型城市服务的房屋的可能性降低了 45.4%。此外，，与其他人相比，拥有更多孩子和更少学习年限的 UPT 支付的可能性也更小。该研究得出结论，边缘化和区域特性是 UPT 支付倾向的主要决定因素，并建议税收政策以确保更多的累进性并减少区域差异。