Evaluating a House Price Index Based on the Sale Price Appraisal Ratio (spar) Method

This paper contributes to the literature on the sale price appraisal ratio (SPAR) as an alternative house price index methodology. Housing market transaction data for 12 cities or districts in New Zealand for the period 1994 – 2004 is utilised to test the effect of measurement errors in assessed values on index accuracy and temporal aggregation on index stability. The main findings are: 1) temporal aggregation effect on the SPAR index stability is a constraint at the monthly level with respect to this New Zealand data set. 2) reporting sample size has a much larger effect on the precision of the SPAR index than the frequency of reassessments. Random measurement errors in assessed values tend to cancel out each other when the number of sales per period is large. 3) Any reassessment tends to disturb the SPAR index for the reassessment period. Consistency errors in assessed values tend to be more likely to accumulate when more frequent reassessment exercises are engaged.