For Discussion: Summary of proposals to amend State Board of Equalization sales ratio calculations June 3, 2010 One of the primary purposes of the sales ratio study is to measure how well assessors track the real estate market when valuing property. Currently the study uses sales from the 12-month period of October through the following September and compares sales prices with the assessors values established as of the January during the period. The sales prices are adjusted to remove non-real estate factors that are not taxable. The sales ratio is calculated by dividing the assessors values by the adjusted sales prices. Median ratios are calculated in all cities, townships, and counties by property type for use by the State Board of Equalization and the Minnesota Tax Court. The Tax Court hears cases based on the assessment year within the study period, so no further adjustments are needed. The State Board of Equalization uses sales ratios that are adjusted to reflect assessors actions to estimate the level of assessment in the next assessment year. The adjustment process is to multiply the median sales ratio by the aggregate percent change from the current year to the preliminary value for next year that the assessor applied to existing property in the jurisdiction. The resulting ratio known as the mini-adjusted ratio must be between 90% and 105% for all property types that have an adequate sales sample in the jurisdiction. Some assessors have expressed concern about the process since the mini-adjusted ratio is often not a precise predictor of the median ratio calculated in the following ratio study. Since the following year study is based on a different sales samples, some difference can be expected. A more serious concern is that in inflationary markets the following year s ratios are often below the 90% threshold while the mini-adjusted ratio was within the acceptable range. The reverse is true in a deflationary market, where the next year s ratios can lie above 105% even though the adjusted ratio was within the range. Sales Ratio: Alternative Calculation Methods Using 2009 and 2010 as examples, the following explains the differing methods for ratio calculations as a form of evaluating the accuracy of assessors market values: The current methodology begins with the simple calculation of the ratio of assessed market value to sales value for the study year: 1
2009 Tax Court Ratio: 2009 Estimated Market Value Sale Price Adjusted to January 2009 This measurement provides an accurate representation of assessment quality for the study period. Unfortunately, this does not reflect current market conditions, it merely reflects on performance in the past. In order to make judgments on changes needed for the current year so that assessed values in the present mirror actual market conditions, the ratio must be adjusted to reflect changes that have taken place since the previous assessment date. The 2010 Tax Court ratio is not yet available though when the State Board of Equalization meets as the sales period has not ended, so current performance must be estimated using past data. Historically this has been done by applying local effort to the ratio of the study period: Mini Adjusted Ratio: 2009 Estimated Market Value X Local Effort Sale Price Adjusted to January 2009 By applying this change to the numerator the local effort acts to estimate what the median sales ratio is in 2010 given changes in assessment level from January 2009. There are a few problems with this method. First of all, local effort is not an accurate indicator of changes in median assessed values. Local effort is derived from the aggregate change in assessed market value from 2009 to 2010, and this only translates into an equivalent change in the median ratio if the changes are uniform across the sample. The 2007 IAAO Standard on Ratio Studies only recommends using the local effort calculation if sales chasing is suspected. Also, in light of recent changes in property class definitions, there is no accurate measure of local effort available for agricultural and, to a lesser extent, residential properties. These shortcomings in the local effort calculation can be addressed by instead using the 2010 assessed values as follows: Preliminary Market Value Adjusted Ratio: 2010 Estimated Market Value Sale Price Adjusted to January 2009 2
This method simply recalculates the median ratio using the 2010 assessed value under the same principles as the local effort adjustment. The same sales sample is used, so the actual adjusted median can be found instead of an estimate based on aggregate changes in value. This requires that updated market value by parcel data be available so that sold properties can be matched by property ID to the 2010 estimated market value. Doing so allows agricultural parcels to be matched regardless of whether updated property codes are used. It also allows the filtering of properties that have had major class shifts. This method does potentially increase the risk of sales chasing by tying the ratio being tested directly to sales from the prior study period. Approximating Current Year Ratios While this change accounts for previously identified shortcomings in using local effort, it does not solve problems relating to time adjustments. Under current methodologies, 2010 assessed values are compared to 2009 market values in evaluating the accuracy of assessments. This creates a lag problem. Assessed values can be outside of the acceptable range in 2009 so long as they adjust assessed values for the following year in a manner that would bring the median ratio back into the range. The problem then arises that the assessed values do not necessarily reflect those of the year that the study covers. Comparing the 2009 adjusted market values from sales to the assessed value a full year later does not account for changes in the actual market during that year. If market values are increasing but changes in assessed values from 2009 to 2010 only bring ratios into the very bottom of the acceptable range, the actual median ratio for 2010 will be below the targeted range even though the adjusted ratio met standards. For example, if the 2009 study ratio for a jurisdiction is at 85% and the assessor applies a local effort of 10%, the resulting mini-adjusted ratio will be 93.5%. But if inflation that year is also 10%, the actual study ratio for the 2010 Tax Court will only be 85%, assuming a representative sample both years. If inflation was actually 15%, the 2010 ratio actually falls to 81.3% despite the 2009 adjusted ratio falling within the acceptable range. Assessors will be a year behind in their valuations and the sales ratio can fail to reflect current market conditions. Also, assessors are limited in how much they can change values using the current method by the range of acceptable median values. As it stands, the 2009 and 2010 Tax Court ratios and the adjusted 2010 median must fall within the 90 to 105 range. If, for example, the 2009 median is at 95%, and the assessor suspects values are changing, he or she can only drop values 5.3% or raise them 10.5% while still staying within the range. But the market may be changing by more than this. This is not the result of an error by the assessor, but rather a shortcoming in the methodology that may prevent justifiable value changes from being made due to constraints in the range and the inability to incorporate changes in the market into the ratio calculation formula. As another example, if the 2009 ratio is below 90% and market values are falling, an assessor must actually increase assessed values to keep the adjusted ratio in the acceptable range, further distorting the ratio for the following year. Instead of reflecting accurate market values, the methodology can lead assessors to aim for inaccurate ratios. 3
These shortcomings can potentially be addressed by incorporating 2010 adjusted market values through a forward time adjustment to the assessment date of the year being analyzed: Regional Representative Proposal: 2010 Estimated Market Value Sale Price Adjusted to January 2010 Under this method, market prices reflect conditions in the actual year the assessment takes place. Ratios then are reflective of assessment accuracy for the current year and current market conditions. The lag is eliminated. This is also more consistent with the language of the IAAO standards which suggest that all relevant factors be adjusted to the same date. In the absence of updated data at the parcel level, the traditional local effort adjustment can be used to estimate the same median value, though with the same pitfalls as before: Minneapolis Proposal: 2009 Estimated Market Value X Local Effort Sale Price Adjusted to January 2010 The IAAO standard recommends that the methodology of the Minneapolis Proposal be used where sales chasing is suspected. As a point of comparison, the Regional Representative Proposal better estimates the 2010 Tax Court Ratio below if the adjusted sale prices are comparable and the time adjustment accurate. The difference in the ratios comes from the samples; the Regional Representative Proposal uses the 2009 sales adjusted forward to January 2010 for its calculation whereas the 2010 Tax Court ratio uses the 2010 sales adjusted to the same date. The differing samples will produce slightly different ratios. The degree of this difference depends on the size and representativeness of the sale samples and the accuracy of the time adjustment. 2010 Tax Court Ratio: 2010 Estimated Market Value Sale Price Adjusted to January 2010 4
Comments: The purpose for adjusting the sales ratio by local effort is to approximate a level of assessment for the next assessment year before sales data has been gathered. The proposed changes seek to improve on this estimate and bring the State Board ratios more in line with the Tax Court ratios. If valid time adjustments can be calculated that accurately reflect changes in market conditions, the estimated ratios for the following year will approximately equal the actual ratios used for Tax Court, assuming a representative sample. The only difference in the ratios will be the difference between the time adjustment and the changes in values made by the assessor. The use of medians derived from sales data adjusted for time as opposed to the local effort calculation is the preferred method suggested by the IAAO standard. Though the details of estimating ratios to judge current market appraisal performance are not laid out in entirety by the IAAO, the use of forward time adjustments to the date of assessment is suggested. Nowhere in the standard is there any mention of using sales values from one assessment year with assessed values from another to calculate ratios, as is done in the current methodology. Alan Dornfest, one of the main sources for the IAAO standard, does explain in more detail the methodology to be used for judging current appraisal performance in the Idaho Ratio Study Manual. He identifies forward time adjustments as the proper methodology and utilizes the method of the Regional Representative Proposal in his Idaho study, with some minor variation. With any ratio study where sales occur before the assessment date, vigilant and consistent sales chasing detection techniques need to be in place and well enforced. In cases where an assessor believes that a time adjustment is not reflective of a local market, sufficient documentation must be provided to a Regional Representative 5
Time Adjustments The utilization of the time adjustment in median ratio calculations does present some problems. A sharp turn in the market at or past the end of the study period will lead to an adjusted market value that is not reflective of current market conditions. A heavier burden is also placed on the accuracy of the time adjustment, a measure that is difficult to produce and rely on given small sample sizes and changing market conditions that cannot be captured by a linear trend line. For the 2009 study, the state was divided into 312 time adjustment region parts when all property types and on-water off-water were considered. Of these, 184 had significant time trends at the 90% confidence level. Many of the regions have small sample sizes and may need to be combined if the sales are to be adjusted forward to the next assessment. 86 counties submitted 2010 preliminary market value by parcel files. Changing time adjustment forward to current assessment year: More consistent with IAAO standards than current local effort calculation Attempts to reflect current market Should increase consistency between tax court ratios and state board ratios May position assessors better for next market turn. If ratios are closer to 100% prior to a turn, assessors will have more room to adjust values and remain within the acceptable range Time adjustments are often not significant in many areas of the state Transition to forward adjustment will be difficult in areas with accelerating markets (e.g. farm areas); would in essence require adding an extra year of market inflation Forces assessors to keep up with rapidly accelerating markets May require additional data collection (Regional Representative proposal requires an early submission of a parcel market value file) 6
Time adjustment procedures available: The current methodology The data is available Does not reflect current market The current methodology, but extend the trend line forward to the next assessment year The data is available Misses late turning points Better reflects current market Impact of any inaccurate time adjustment magnified Trending of last 3 months of the study period along with any early sales from the next study period Measures time trend in next study period Spotty data from incomplete sale submission Better reflects current market Small sample in almost all areas Add late sales as available and use 12 month calendar year trend Develops a calendar year trend into the Spotty data from incomplete sale next study period submission Better reflects current market Uniformity limited to data availability Minnesota Department of Revenue Sales Ratio Unit 7