Decision rendered September 6, 1996.

CARL N. BYERS, Judge.

Plaintiff (taxpayer) appeals the 1994-95 assessed value of his personal residence, claiming the benefit of an adjudicated value under ORS 309.115 rather than contesting the real market value under ORS 308.205. The 1994-95 assessed value was increased 30 percent over the prior year's assessed value. Taxpayer claims that the 30 percent increase in one year is based on a defective ratio study and his assessed value should be reduced accordingly. The Department of Revenue (department) defended the ratio study and the resulting trending factor.

The subject property consists of three tax lots with taxpayer's home sitting on one of the lots. The home was constructed in 1988 and fronts on the Clackamas River. Taxpayer appealed the real market value of the property in prior years and obtained an adjudicated value. For the year in question, the adjudicated value was trended up 30 percent based on a ratio study comprised of only five sales. ORS 309.115 permits taxpayers to claim the benefit of a prior adjudicated value for up to five years with certain exceptions. One of those exceptions is:

"(2)(b) Annual trending or indexing applied to all properties of the same property class in the county, or within clearly defined areas of the county under ORS chapter 309."

ISSUE

The issue before the court is whether the trending factor applied to the subject property under ORS 309.115 was correct.

ANALYSIS

A brief explanation of ratio studies is necessary to understand the following analysis. Due to limitations of time and money, assessors are unable to physically reappraise every property every year. To comply with the statutory mandate that properties be assessed at their real market value every year, assessors apply a trend factor to increase or decrease assessed values to reflect trends in inflation or deflation. The trends are determined by ratio studies. A ratio study is a statistical analysis of the relationship between the sale prices of properties and their assessed values. Accuracy of the study depends upon classification of the properties, whether available sales are representative of the class, and a number of other factors. The department published an Assessor's Ratio Procedures Manual, 1994, which sets forth the standards and methods to be followed.

Taxpayer challenges the county's ratio study on several grounds. He primarily argues that the study was too small and did not properly represent the properties trended. Taxpayer believes one of the five sales in the study, 855 Edgewater, had a carport and a driveway added after the assessed value for 1993-94 was set. If true, the 1994 sale price of the property reflected, in part, improvements made to the property, not inflation. However, the assessor's records show that the paving and carport were added to the 1991-92 tax records in the amount of $5,580. Consequently, the value of those improvements would have been included in the 1993-94 assessed value.

Taxpayer also believes the assessor's records showed one of the sale properties had only one story containing 1,646 square feet when in fact it was a two-story property containing 2,788 square feet. He also contends that the sale included other property. However, the appraiser testified that the other property was only .04 of an acre, constituted part of the same riverfront homesite, and was included in the assessed value. He also testified that the assessor's appraisal card shows that the second floor was checked and valued. Apparently, taxpayer was misled because older computer printouts reflected square footage for only the main floors rather than living areas.

The county excluded the sale of three five-acre lots because they constituted a sale by the federal government to a local government unit. Taxpayer testified that the three governmental sales were listed on the market and their sales prices represent market value. The department's witness automatically excluded these sales from the ratio study and consequently did not analyze them to determine whether they constituted market transactions. His exclusion of these sales is not consistent with the department's published manual. The department's witness, although acknowledging that the sales "may be market," believed that the manual required automatic rejection based on the rejection codes listed on page 7-3. However, this listing merely suggests codes to use as reasons when sales are rejected. The manual does not require automatic rejection of such sales.

1, 2. The manual states: "The ratio study must include all sales which are free-market transactions between a willing buyer and a willing seller." (Emphasis added.) It provides that assessors should reject sales if the sales are not good evidence of fair market value. The manual lists certain sales that do not require a sales record, e.g., if the grantor is a sheriff, receiver, or other court officer disposing of property under a judicial order or administrative proceeding. However, it is clear that assessors should analyze all sales to determine whether they constitute evidence of market value. The manual expressly states:

"Transfers by government agencies, lending, and financial institutions may have significant influence in the real estate market. Sales by these grantors must be confirmed to establish whether the sale price is indicative of market or is a liquidation price, and the property's physical condition is the same as when it was assessed. Depending on the results of confirmation, these sales may be usable in the sales listing or in a supplemental study."

Therefore, the assessor's office erred in automatically excluding the government sales rather than determining whether the sales were market transactions.

Taxpayer also asserts that sale number five on exhibit 16 should have been included in the ratio study. The department's witness agreed and prepared a new study to include this sale. Including that one sale changed the range of trend factors. The witness explained that in performing a ratio study, if a property's assessed value is only 74 percent of its sale price, the 74 percent is divided into one to determine the trend factor, e.g., 1 ÷ .74 = 1.35 or 35 percent trend. As required by the manual, the witness calculated both the mean, the weighted mean, and the median of the sales ratios to determine the range of potential increase in value. The ratio study used in trending the subject property up 30 percent indicated a range of 35 percent to 54 percent. When sale number five is included, the range drops to 28 to 37 percent, causing the department's witness to revise his opinion and indicate that the appropriate trend factor for the subject property should have been 28 percent rather than 30 percent.

The department's witness also performed a sales ratio study that straddled the assessment date. Normally, sales for ratio studies are drawn during the year prior to the assessment date. The straddle ratio study included sales before and after the assessment date. The witness performed this ratio study using three different time periods: 8 months, 16 months, and 2 years. These studies gave an indicated range of trending factors as follows:

INDICATED RANGE OF FACTORS

EIGHT-MONTH STUDY 16-MONTH STUDY TWO-YEAR STUDY _________________ ______________ ______________ 1.23 to 1.30 1.28 to 1.43 1.27 to 1.30

The department's ratio study manual states that the "price-related differential" (the mean ratio divided by the weighted mean ratio) gives some indication of whether higher price properties are overassessed or underassessed. The ratio study in question had a mean of .78 divided by a weighted mean ratio of .73, for a price-related differential of 1.07. This would indicate that higher priced properties are under assessed. However, it is outside what the manual considers an acceptable range. The manual states that when sample sizes are small, as in this study, "price-related differentials outside the acceptable range may occur simply because of random sampling error." The department's straddle study yields price-related differentials within the acceptable range, which suggests that the price-related differential of the actual ratio study is an aberration, due to a random sampling error. The court recognizes that the department's straddle study presumably also automatically excludes government sales and incorporates the same types of errors made in the original ratio study. Consequently, its reliability is subject to question, although to a lesser degree.

Based on the above, the court concludes that the ratio study used to trend the subject property 30 percent for the 1994-95 tax year was defective. The department's witness conceded that 30 percent was an error and believes the correct trend should be 28 percent. The court does not accept taxpayer's position that the trend is only three to five percent per year. Based on the straddle study, judgment will be entered reducing the assessed value of the subject property to an amount that equals the prior year assessed value trended 23 percent. Costs to neither party.