Assessing the Viability of Simplified Automated Valuation Methodologies for Property Taxation: Evidence from Pakistan

Status

Upcoming

Recent research has highlighted the potential of simplified valuation methodologies for improving the accuracy, transparency and cost-effectiveness of property valuation exercises, particularly in lower income country contexts with limited administrative resources, limited real estate market information and high risks of corruption. Reliance on simplified valuation methodologies implies a belief that the cost savings, added transparency, and produced subjectivity and risk of corruption will outweigh any loss of precision. This paper investigates those claims empirically by piloting the use of a simplified “points based” valuation approach in Punjab, Pakistan. We first seek to estimate the accuracy of traditional expert driven valuation approaches by sending multiple valuers to the same property, in order to estimate the extent of subjectivity, and by comparing a sample of existing administrative data to actual market values, in order to estimate existing human error and/or corruption. We then develop a simplified formula for estimating property values from easily observable property characteristics, and explore both the overall fit of those simplified models and their sensitivity to the inclusion or exclusion of additional information.