Conference Agenda

01-08: Methodological approaches to urban property valuation
Tuesday, 26/Mar/2019:
8:30am - 10:00am

Session Chair: Ruud Kathmann, Netherlands Council for Real Estate Assessment, Netherlands, Netherlands, The
Location: MC 8-100


Self-declaration of value: an option for the urban property tax

William McCluskey1, Riel Franzsen1, Peadar Davis2

1African Tax Institute, University of Pretoria, Pretoria, South Africa; 2School of the Built Environment, University of Ulster, N. Ireland, UK

The ad valorem property tax is a presumptive tax based on an estimate of the property’s value. The estimation of value is normally undertaken by the valuation authority. To assist the authority, there is usually a legislative requirement whereby the taxpayer (whether owner or occupier) is obliged to declare certain information regarding the property. The information sought is typically related quantitative characteristics of the property such as age, size, and accommodation. In addition, some value based information such as tenancies, lease details, business/trade turnover and rents can be sought. This is information that the valuation authority use in their estimation of the value of the property.

The focus of this paper is to consider whether the owner should be obliged to submit a declaration of their property’s value. The question is whether these valuations could be used for the property tax.

Valuing property with bad data: utilizing GIS and spatial modeling to achieve equitable property tax valuations in the face of incomplete data

Paul Bidanset1, Jones Brent2

1IAAO; 2Esri

Missing, incomplete, or inaccurate data have the ability to compromise any predictive model (Beaver, 1966; Pifer & Meyer, 1970; Martin, 1977; Altman, 1981; Bansal et al. 1993). For models used for ad valorem property tax purposes, such data inadequacies can result in financial burdens that arise from inaccurate valuations. Recent advances in spatial modeling methodologies and the availability of open data sources have presented governments with ways to “do more with less” – specifically achieving more accurate valuations without incurring additional data collection costs.

This research will present recent findings on how spatial modeling and open data can be harnessed by governments to promote more fair and uniform property valuations with fewer costs incurred. This presentation will bridge the current literature gap by making explicit methodological prescriptions for valuations that will be yield uniform and equitable valuation results for governments faced with technological or financial constraints.

Response surface analysis (RSA): modeling values in geographically sparse markets

William Mccluskey1, Paul Bidanset2, Peadar Davis3, Michael McCord3

1African Tax Institute, University of Pretoria, Pretoria, South Africa; 2International Association of Assessing Officers, Kansas, United States; 3School of the Built Environment, University of Ulster, N. Ireland, UK

Due to physical, legal, and other barriers, as well as cost-prohibitive reasons associated with data collection and storage, sparse data can be a common hurdle in the effectiveness of governments who depend on or are considering the implementation of a property tax regime. The ability to estimate accurate, equitable property valuations is oftentimes a difficult task, particularly in areas with little or no sales transactions. In developing and transitioning economies with limited, inaccurate, or outdated sales information, the data needed to create reliable estimates of value is often times very difficult, or even impossible, to attain.

Standard price points in spatial interpolation. A case study

Risto Peltola, Mikko Korpela, Pauliina Krigsholm, Arthur Kreivi

National Land Survey, Finland

This paper explores various spatial interpolation techniques and tries to find an optimal mix of automa-tion and manual fine tuning for mass valuation purposes for property taxation. The value of land should be estimated to the tax base of 2 million units, in a system of higher tax rate on land than on buildings. Machine learning techniques such as ordinary kriging, empirical bayesian kriging (EBK), geographically weighted regression (GWR), inverse distance weighting (IDW) and spatially constrained cluster analysis (SCCA) have been tested. The paper offers a comparison between those methods and a compari-son to a more manual approach.

Using remote sensing data and machine learning to value property in Kigali, Rwanda

Felix Bachofer1, Jonathan Bower2, Andreas Braun3, Paul Brimble4, Patrick McSharry5

1German Aerospace Center; 2International Growth Centre, Rwanda; 3University of Tübingen; 4Ministry of Economic and Financial Planning, Rwanda; 5Carnegie Mellon University

This paper develops two property valuation models for Kigali, Rwanda, and tests them on a unique dataset combining remote sensing data for buildings in Kigali, with sales transaction data for 2015. This paper credits and builds on a similar paper by Deininger et al (2018) but also covers both the built up area of Kigali and the whole of Kigali Province, it addresses temporal prediction issues beyond 2015, it employs an expanded set of variables, and it uses machine learning techniques to employ Maximum Relevance Minimum Redundancy to select the model that best predicts property price data using Ordinary Least Squares. The model in this paper is intended as a prototype of a Computer Assisted Mass Appraisal system for Kigali that could be used to calculate the revenue potential of a new property tax introduced in Rwanda in 2019, and to help detect under-declaration of property values for tax purposes.