Computer assisted valuation of land and property
Land or property tax is one of the oldest forms of taxation; it is relatively fair and it is not easy to hide properties. Given that real estate is inherently local, revenues from this tax are generally levied locally. Often it is a key tax for the local governments to finance urban services and amenities. Relating the height of the tax to the market value of the property (land a/o buildings) is considered the most equal approach. But only a rather small portion of the properties is sold every year, so we need to come with an estimate of value in another way. Computer Aided Mass Appraisal (CAMA) is an approach to do so by analyzing how the various characteristics of the sold properties influenced their market price. In this way, it is possible to generate an empirical model to estimate the potential market value for the rest of the properties, by analyzing their structural and locational characteristics.
The study will look into different methods to implement CAMA. This can focus on the models and statistical approaches to compute the value for the non-sold properties. It can also focus on determining which characteristics should be used to best describe the properties for this purpose, and/or to study where to find the information regarding those characteristics. This could be re-use of existing (geo) data sets, but also call for collecting data in novel ways.
The topic will bring together a feeling for property value, and thus the societal importance of properties, with more technical skills like data collection, data integration, some base programming and modelling. As these approaches can also be seen as applying algorithms, depending on the precise issue at hand AI and spatial models can come into play.
Where possible, connection will be sought to a real ongoing property taxation project via contacts with international organizations. Where needed, input will be sought from experts outside ITC, both within the UT and with relevant agencies.