ON THE MEASUREMENT OF ASSESSMENT ERROR

Differences market value ratios in assessed as measured value/ by market value r tios as measured by the coefficient of dispersion or some other statistic for the same class of property in the same taxing district are frequently used as a measurement of "assessment error." The first argument of this paper is that this measure is inaccurate because assessment errors and market prices are not necessarily independent of each other. It is also frequently alleged that differences in assessment/market price ratios for different classes of property also represent "assessment error' or "assessment bias." The second argument of this paper is that this conclusion also is unwarranted as there is little reason to believe that property taxation has the same effect on property prices for all classes of property.