HB3211 Addresses Issue in Property Value Ratio Study
In any ratio study that relies on property sales, a general question arises whether ratios of appraised value to market value on sold properties accurately reflect the ratios on unsold properties. The Property Tax Division (PTD) faces this issue when it performs the Comptroller's Property Value Study (PVS) to determine the total taxable wealth of each school district for state funding purposes. The PTD also faces this problem when it reviews the work of county appraisal districts (CADs) to determine the level and uniformity of their appraisal work.Ratio studies in general
During the last legislative session, the Texas Legislature passed House Bill 3211 requiring PTD to ensure "that different levels of appraisal on sold and unsold property do not adversely affect the accuracy of the study." PTD met with appraisal districts, taxpayers, and other affected parties in an effort to implement this legislative mandate. Here's a general review of the issue of selective reappraisal.
Ratio studies assume that the typical ratio (weighted mean for instance) computed from a property sample is the "true ratio" for all property in a category. This depends not only on whether each sale is a valid market transaction, but also on whether the appraisal district used similar appraisal procedures to determine final appraised values on sold and unsold properties. The Standard on Ratio Studies adopted by the International Association of Assessing Officers (IAAO) in 1999 states:Effect on study results
"The objective of a ratio study is to determine appraisal performance for the population of properties. As long as sold and unsold parcels are appraised in the same manner, statistics calculated in a sales ratio study can be used to infer appraisal performance for unsold parcels. However, if parcels that sell are selectively reappraised based on their sales prices or some other criterion (such as listing price), sales ratio study uniformity measures will not be valid (appraisals will appear more uniform than they are). In this situation, measures of appraisal level will also be invalid unless similar unsold parcels were reappraised by a method that produces the same percentage of market value (appraisal level) as on the parcels that sold. Assessing officials must ensure that sold and unsold parcels are treated equally."
The term used by the IAAO to describe selective reappraisal of sold property based on sales price or some other criteria is "sales chasing."
The effect of selective reappraisal is addressed in the Property Appraisal and Assessment Administration (PAAA) textbook from IAAO in 1990. The textbook states:Effect on schools and taxpayers
- "In all probability, calculated measures of central tendency [weighted mean ratios, for example] will be artificially high and measures of dispersion [coefficients of dispersion (CODs), for example] will be artificially low." Bracketed language added.
The reason for this supposition is that nationwide, ratios are more commonly low than high, indicating a typical condition in which appraised values fall below market value. This is especially true during periods of rising real estate markets. If appraised values on ratio study sales are adjusted close to sales prices, ratio study results normally reflect an artificially high level of appraisal. At this point, the ratio study would merely represent appraisal conditions in a small subset of the category - namely, the selectively reappraised properties. The biggest problem is that the true level of appraisal for the entire property category becomes clouded and indeterminate. The ratio study, intended to estimate the true level of appraisal, fails in its mission.
If the Comptroller's PVS finds artificially high levels of appraisal, some school districts could receive more funding than they are entitled. Since this unfair result could divert funds that could otherwise be available for equitable distribution to all school districts, some taxpayers pay more in local school taxes since their districts didn't receive a larger share. Taxpayers who purchase property also pay more than their fair share of taxes annually when their property is appraised at a higher level of appraisal than other property.Difference in appraisals
When the typical appraisal level in a category of property is close to market value, it is likely that a ratio study of the category will include sales with ratios that tend to fall fairly evenly on both sides of 1.00 (that is, 100 percent of market value). Such a ratio study concludes that the category level of appraisal is at market value. If few appraised values differ greatly from sales prices, variability as measured by the coefficient of dispersion (COD) is low and the ratio study concludes that appraisal uniformity is good. All of these assertions are valid provided that the property sample in the ratio study is representative of sold and unsold parcels in that category.Solutions
With selective reappraisal, the level of appraisal approaches 1.00 and the COD is very low. How can one distinguish between a result that applies only to the sample and indicates selective reappraisal, and a result that can be applied validly to the category and indicates good appraisal conditions?
While there is not one simple answer to this dilemma, several methods to determine whether sold and unsold properties are being appraised in the same manner are available. Some of these methods are discussed in detail in Appendix 20-2 of Property Appraisal and Assessment Administration (PAAA) (IAAO, 1990) and the Standard on Ratio Studies (IAAO, 1999). Two of these methods are discussed below.
Using distribution of ratios to test for selective reappraisal. PAAA observes the following: "Low CODs (15.0 or less) tend to be associated with good appraisal uniformity."
The text goes on to state: "CODs of less than 5.0 are very rare except in (1) subdivisions in which lot prices are strictly controlled by the developer; (2) extremely homogeneous property groups, such as condominium units all located in the same complex; (3) appraisal ratio studies in which the assessor's values and the independent appraisals reflect the same appraisal manuals and procedures; or (4) appraisals that have been adjusted to match sales prices."
In other words, except in highly constrained or unusual situations, very low CODs are indicators of selective reappraisal. To state this more plainly, when the COD is very low, an unreasonable proportion of appraised values will be very close to sales prices. In many instances, one can look at the distribution of ratios in the sample, count the number of ratios close to 1.00, and determine from the proportion of ratios close to 1.00 whether selective reappraisal is likely.
Mathematically, one cannot determine directly from the COD how many ratios should be expected to fall in a given range. This can be done, however, using another statistic, the standard deviation. By computing the proportion of ratios that would be expected to fall within a particular narrow range of the mean, given the lowest likely standard deviation, it is possible to obtain an indication of the likelihood of selective reappraisal.
For example, with a standard deviation of .05 and given a normal distribution, one would expect to find about 32 percent of the ratios between 98 and 102 percent of market value, given a mean of 100 percent. It is unlikely that a sample with more than this proportion of ratios in such a narrow range is representative of all property in the category.
Using comparison of average value changes to indicate selective reappraisal. Another common method is to compare appraised value changes on sold properties with appraised value changes on unsold properties. In the absence of selective reappraisal, if the sold properties increased by 10 percent in appraised value, then the unsold properties in the category should have increased by about this same amount.
Discrepancies between patterns of change on sold and unsold properties can be compared to see if statistically significant differences exist. If so, selective reappraisal is likely unless the sales sample is not proportionally representative of value influences on the property population. For example, geographic hot spots (areas with higher than average sales activity and value increases) that have been recently reappraised may give the appearance of selective reappraisal in this test if the hot spot is tested as part of a broader population with fewer sales and more moderate appreciation.
Solutions in ratio studies to selective reappraisal may be achieved in several ways. The most commonly used are (1) substitution of independent appraisals and (2) use of sales occurring during a later period.Positive aspects
The PTD already uses appraisals to supplement samples when inadequate numbers of sales are available. Five states in the United States use only appraisals in their ratio studies. Although this is a resource-intensive procedure and often presents additional quality control problems, it is an acceptable alternative that can produce a more representative sample.
Sales are usually considered the best proxies for market value and, therefore, the best indicators of appraisal conditions. In contrasting appraisal-based and sales-based ratio studies, PAAA states: "When enough sales are available, however, sales ratio studies produce the most efficient and objective measures of appraisal performance...."
Since CAD appraised values become essentially "locked in" following final approval by each county's appraisal review board (ARB), use of sales that occur too late in a given year for ARB consideration should prove an effective substitute for a sample with "selected" sales. Sales prices must be adjusted for time, so that the price will reflect value as of the appraisal date (that is, January 1).
Similarly, late sold properties must be reviewed for significant changes to the property that occur between the appraisal date and the sales date. If a property has significant changes, the sales price should be adjusted to remove the effect of the changes or the sale should be excluded. Late sales should also be reviewed to ensure they provide a representative sample.
In addition to the obvious result of more representative ratio studies, several other positive results can be expected if this issue is effectively addressed.Conclusion
In an environment where the PVS results are clouded by selective reappraisal, taxpayers, school districts and taxing units are unable to tell whether they are being treated equitably. If the PTD modifies the PVS to produce a representative sample even when selective reappraisal is evident, taxpayers will receive better information about their level of appraisal and uniformity in comparison to other taxpayers. School districts will be more certain that they are receiving the proper amount of funding. Other taxing units will be assured of appraisal accuracy and uniformity.
Each CAD's goal is to correctly appraise all parcels. The PTD's goal is to do a fair and accurate ratio study to determine appraisal conditions and to correctly apportion state aid to school districts. To accomplish both goals, the PVS requires a valid and representative sample. If an ARB has adjusted the appraised values of a large portion of recently sold properties, later used in a ratio study, then the PTD also must consider the effect of ARB actions in determining whether the study is valid. Ratio study results could be as potentially inaccurate as if the adjustments occurred at another level. Discovery and correction of non-representative samples caused by selective reappraisal will help assure that CADs and ISDs are studied on a level playing field, regardless of ARB decisions.
If selective reappraisal is determined to exist, it is not necessarily a criticism of the CAD or the ARB, both of which are presumed to be making honest efforts to correct perceived inequities and to make similar adjustments on unsold properties. The problem is that the effectiveness of the ratio study is reduced and the equalization effort to ensure proper distribution of payments between state and local governments is jeopardized.
PTD encourages participation in implementing the new law with suggestions, comments and questions. To comment, e-mail Tim Wooten, PTD Area Manager, Property Value Study, at e-mail address email@example.com. Or, call 1-800-252-9121, ext. 5-9838.
Contributing to this article: Alan Dornfest
Note: Alan Dornfest is a nationally recognized property tax expert in ratio studies. He is supervisor of the policy development section of the legal/tax policy division of the Idaho State Tax Commission. Alan has written extensively on ratio studies and was principal author of the International Association of Assessing Officers (IAAO) Standard on the Valuation of Property Affected by Environmental Contamination (1992) and the IAAO Standard on Property Tax Policy (1997). He was a contributing author for the 1990 and 1999 IAAO Standard on Ratio Studies and co-author of the ratio study sections of Property Appraisal and Assessment Administration (IAAO 1990). Alan is chair of the Technical Standards Subcommittee of IAAO's Research and Technology Committee. He has received three IAAO presidential citations for service. Alan also has published numerous articles and made numerous presentations in the property tax assessment field. He has taught ratio study workshops in many states and served as a consultant reviewing ratio study practices in Texas, Washington, Kansas and Illinois.