A simulation model for the comparison of sampling strategies used in estimating total residential market value for a geographic area

  • Authors:
  • James M. Kraushaar

  • Affiliations:
  • -

  • Venue:
  • WSC '76 Proceedings of the 76 Bicentennial conference on Winter simulation
  • Year:
  • 1976

Quantified Score

Hi-index 0.00

Visualization

Abstract

The total property market value for a taxing district is used to: distribute federal, state, and county taxes; apportion joint school district tax levies; limit taxing and borrowing powers. This total market value cannot be calculated by simple summing the individual property values as listed on the local assessment roll because assed values may or may not reflect current market values, depending on many factors related to the assessment process. Consequently, total district assessed value cannot be a basis for tax allocation. Stratified random sampling and ratio estimators are often used to reduce the large sampling error. However, measuring the precision (i.e., bias and variance) of these ratio estimators is difficult when samples are small because the second product moment equation may be unreliable.(1) Without a measure of precision, the effectiveness of various sampling design factors (e.g., number of strata, stratum boundary rule, and type of ratio estimator) can not be evaluated. A simulation model composed of three separate modules was developed to measure the precision of the total market value estimators for a population of properties.