| There are two policy issues guiding this research: horizontal equity in property tax burdens, and the adequacy and predictability of the revenue source (Netzer 1966; Ladd and Yinger 1989; Rosen 1992). Previous empirical studies of assessment performance focused their attention on socio-economic and administrative variables, failing to consider explicitly the political factors that are discussed in the theoretical literature. A comprehensive study of all these factors permits to address policy issues related to property value assessment administration.; This research adopts an open model of the organization involved in property value assessment including socio-economic, administrative, and political. A logarithmic multiple regression analysis is performed using cross-sectional data {dollar}rm (n=433){dollar} of assessing jurisdictions in the United States for 1982, 1992, and 1982-92. The value added of this research is contained in the comprehensive theoretical model that is used for empirical analysis of relevant data.; From the empirical results, there are a seven general patterns regarding property value assessment performance in the USA during the 1982-92 period. First, notice that a comprehensive model (Model 2) of property value assessment performance is more powerful in explaining Residential and Total variance in assessment accuracy than a traditional model (Model 1) which only contains socio-economic and administrative variables.; Pointing to the partial empirical results, there are six other general patterns (See Table 13). First, there is better performance in rural relative to urban jurisdictions. Second, assessment jurisdictions with relatively higher Constant-Dollar Administrative Expenditures seem to have better assessment performance. Third, elected local CAO's reduce assessment dispersion more relative to appointed local CAO's. Fourth, the more value exempted from property taxation, the greater the assessment dispersion in both 1982 and 1992. Fifth, there is weak empirical support for the hypothesis that increasing centralization reduces aggregate variation in local assessments. Finally, there is strong empirical support for the negative and statistically significant effect of Assessment Level on assessment variation in both annual and growth ratio studies. Thus, an increase in the assessment-to-sales ratio for the entire jurisdiction is perhaps the most important factor in bringing down variation in initial local assessments and improving assessment performance.; In conclusion, five major policy implications are derived from this study of assessment performance in the USA: (1) increase Assessment Level as close to 100% of market value as feasible; (2) reduce the annual level of exemptions from property taxation; (3) augment centralization of initial assessment tasks by collaboration between state and local government agencies involved in property tax administration, financial management, and economic development; (4) support the election of local chief assessing officers (CAO's); and (5) increase constant-dollar administrative expenditures in areas such as assessor's training, and computerization. |
