The property tax system that the Tanzania Revenue Authority (TRA) inherited has a low proportion of total properties registered and an even smaller proportion valuated. It uses a valuation method that is regressive, expensive and significantly undervalues properties, and the effective tax rates are minuscule once true property value is taken into account.
A flat tax system could be a tool to expand the tax net radically and quickly, but there are serious challenges about ensuring equity and good revenue generation. Flat tax rates are applied in a number of countries (e.g. Baltic countries, USA), but a system of flat-tax – when the same tax amount is attached to all properties – has not attracted the attention of academia or the practitioners, because of rare appearance like the current rules in Tanzania.
The overarching objective of this study on property tax in Tanzania is to explore options, and test modalities and conditions for a banded flat-tax system for taxing real properties in a fair, efficient, and effective manner in Tanzania. The single levy that the Government of Tanzania has been considering is unprecedented in international comparison and it may reduce, instead of increase, tax revenues. However, testing a system of flat amount calibrated to the economic structure and characteristics of taxable properties and regions and cities is worth testing and may offer options and modalities for cost-effective increase of the tax base. The study is expected to inform TRA and the broader Government of Tanzania’s redesign of the property tax system, which is imminent.
TRA administrative data of the existing property tax system provides a rich factual base for analysis; but this study will require special structuring and cleaning of TRA data as well as supplementation with a moderate volume of new data too, which the project team will collect, alongside with TRA officers. The team will define clusters of properties relevant to consistently map the universe of properties existing in Tanzania. This will be done such that properties in each cluster are sufficiently similar for taxation purposes and therefore can be considered as a homogeneous group, similar to the banded approach used for council tax in the UK. The data analysis and model testing will follow simplified mass valuation methodologies to ensure statistically verifiable and cost-effective definition of clusters, and estimation of comparative property values to serve as a basis for defining flat tax rate for each cluster.