BISON TOWNSHIP, Plateau Township, Lincoln Township, Rainbow Township, Lone Tree Township, Scotch Cap Township, Wilson Township, Vickers Township, Strool Township, Maltby Township, Viking Township, and Horsecreek Township, Appellants, v. PERKINS COUNTY, South Dakota, Appellee, 607 N.W.2d 589

MILLER, Chief Justice. [¶ 1.] Eleven townships in Perkins County appeal their 1998 real estate assessments. Because County included non-arms-length transactions in its market analysis and failed to properly re-evaluate the use of assessment zones, we reverse and remand. FACTS [¶ 2.] In 1993 Perkins County (County) exercised its statutory right to create “zones” of different assessment areas within the county. Where there existed an identifiable region that deviated more than 10 percent in value from the county average, a separate assessment zone was created. As a result, Perkins County is divided into four zones.

Appellants (Townships) own property located in Zones 1 and 3. [¶ 3.] When assessing property, County relies mainly on two factors: soil samples and market values. It utilizes soil surveys prepared by the U.S. Department of Agriculture, which categorize soil into eight types. The most productive soil is assigned a ratio of 1. For each township, an average soil rating is determined based upon the composition of soils in that particular area.

For example, a township may have an average soil rating of .654 (with 1 being ideal), depending upon the capability of the soil to produce. [¶ 4.] County also conducts a market analysis using recent sales of agricultural land. Sales data is reviewed to determine the “top dollar value” (per acre) for each zone. The top dollar value in each zone is assigned to soil with a ratio of 1, the most productive soil. For the 1998 assessment year, Zones 1 and 3 had an assigned top dollar value of $256.25. Zones 2 and 4 had a top dollar value of $205.00.

Zones 1 and 3 have higher top dollar values because County has determined that those areas deviated more than 10 percent in value from the county average. [¶ 5.] After compiling soil sample and market value information, the average soil rating for each township was multiplied by the top dollar value in that zone. As a result of differing top dollar values, townships that had nearly identical average soil ratings had different assessment values. For example, DeWitt Township, which is located in Zone 1, had an average soil rating of .469.

When multiplied by the top dollar value of $256.25, the average assessed value was $120.18 per acre. On the other hand, Wells Township, which is located in Zone 4, had an average soil rating of .468. When this ratio was multiplied by the top dollar value of $205.00, the average assessed value was $95.94 per acre. [¶ 6.] Townships appealed their 1998 assessments based upon these apparently disparate results. They challenged whether the assessment zones utilized by County were accurate, and whether the County used only arms-length transactions in conducting its market analysis of sales in the county.

The Office of Hearing Examiners (OHE) found that County did not perform an analysis for 1998 to determine whether the assessment zones were accurate. OHE further found that County did not use arms-length transactions in its market analysis. OHE therefore ordered Townships’ assessments be reversed; however, it directed that the top dollar value of Zones 2 and 4 be applied to Zones 1 and 3. County appealed OHE’s decision to the circuit court.

The court reversed OHE’s decision, stating in its findings of fact and conclusions of law: The court finds as a fact that the Director of Equalization of Perkins County correctly performed his duties pursuant to statute and the rules and regulations of the State of South Dakota Department of Revenue; The Court finds as a fact that the appellees failed to overcome the presumption that the tax official did his duty in accordance with the law and that his figures were correct; The Court finds as a fact that the appellees failed to overcome the presumption that the tax official did not act unfairly and arbitrarily regarding the assessment of property in Perkins County; The Court finds as a fact that the Hearing Examiner was clearly erroneous as a matter of law in her findings and conclusions.

The Hearing Examiner was clearly erroneous as a matter of law in light of the entire evidence and record hereinf.] [¶ 7.] Townships appeal the circuit court’s decision, raising the following issues: 1. Whether County violated SDCL 10-11-56 by not using arms-length transactions for comparable sales. 2. Whether County violated SDCL 10-6-33.6 by failing to properly re-evaluate the use of assessment zones. STANDARD OF REVIEW [¶ 8.] As we recently stated in Butte County v.

Vallery, 1999 SD 142, ¶ 8, 602 N.W.2d 284, 286-87: This is an appeal of a tax assessment pursuant to SDCL 10-11^13 and thus it is proeedurally governed by SDCL ch 1-26. Under SDCL 10-11-42.1, the hearing examiner tries the issues de novo. On appeal both the circuit court and this Court review that decision as set forth in SDCL 1-26-36. This standard of review requires us to accord great weight to the findings and inferences made by the hearing examiner on factual questions. Clarkson & Co. v. Harding County, 1998 SD 74, ¶ 5, 581 N.W.2d 499, 501 (citing Sopko v.

C & R Transfer Co., Inc., 1998 SD 8, ¶ 6, 575 N.W.2d 225, 228). “When the issue is a question of fact, we ascertain whether the administrative agency was clearly erroneous.” Loyal Order of Moose Lodge v. Pennington County, 1997 SD 80, ¶ 5, 566 N.W.2d 132, 133. (Citations omitted.) See also West Two Rivers Ranch v. Pennington County, 1996 SD 70, ¶ 6, 549 N.W.2d 683, 685 (quoting Lincoln Township, 1996 SD 13, ¶ 24, 543 N.W.2d at 259). “When the issue is a question of law, the decisions of the administrative agency and the circuit court are fully reviewable.” Vallery, 1999 SD 142, ¶ 8, 602 N.W.2d at 287.

DECISION [¶ 9.] 1. County violated SDCL 10-11-56 by not using arms-length transactions for comparable sales. [¶ 10.] All property shall be assessed at its true and full value in money. SDCL 10-6-33. However, exact uniformity and mathematical accuracy in valuations are impossible. Kindsfater v. Butte County, 458 N.W.2d 347, 350 (S.D.1990) (citations omitted). Therefore, uniformity and equality of taxation is preferred over the standard of full and true value when both objectives cannot be secured.

Id. [¶ 11.] There is a presumption that tax officials act in accordance with the law and not arbitrarily or unfairly when assessing property. Amert v. Lake County Bd. of Equalization, 1998 SD 66, ¶ 14, 580 N.W.2d 616, 618-19 (citing Richter Enterprises, Inc., v. Sully County, 1997 SD 61, ¶ 7, 563 N.W.2d 841, 843) (other citations omitted). The taxpayer also has the burden of overcoming the presumption that the assessment is correct by producing sufficient evidence to show the assessment was in excess of true and full value, lacked uniformity in the same class, or was discriminatory.

Id. [¶ 12.] OHE found that County used non-arms-length sales in its market analysis, because it included intra-family real estate transfers. In addition, OHE found that County failed to use arms-length sales in its market analysis, even though such data was available. After reviewing the OHE hearing transcript, exhibits, and records, and hearing oral arguments from the parties, the trial court reversed OHE’s findings. However, the court’s findings only generally state that Townships did not overcome the presumptions of correctness and fairness in the assessment process.

Mindful of our standard of review, we hold the trial court erred because there is evidence that County used prohibited non-arms-length transactions in its market analysis. [¶ 13.] SDCL 10-6-33 concerns the procedure to be utilized in making property valuations: All property shall be assessed at its true and full value in money. The true and full value is the taxable value of such property upon which the levy shall be made and applied and the taxes computed.

In determining the true and full value of property the director of equalization may not adopt a lower or different standard of value because it is to serve as a basis of taxation. The director may not adopt as a criterion of value the price for which the property would sell at a forced sale, or in the aggregate with all the property in the third class municipality or district. The director shall value each

article or description by itself and at an amount or price as he believes the property to be fairly worth in money. The true and full value shall be determined by appropriate consideration of the cost approach, the market approach and the income approach to appraisal.

The director of equalization shall consider and document all elements of such approaches that are applicable prior to a determination of true and full value. [¶ 14.] The assessment of agricultural land is guided by SDCL 10-6-33.1: The true and full value in money of agricultural land, as defined by § 10-6-31, which has been in primarily agricultural use for at least five successive years immediately preceding' the tax year for which assessment is to be made shall be the market value as determined for each county through the use of all comparable sales of agricultural land based on consideration of the following factors:

(1) The capacity of the land to produce agricultural products as defined in § 10-6-33.2; and

(2) The soil, terrain and topographical condition of the property including but not limited to capability, the land’s use, climate, accessibility and surface obstructions which can be documented through an analysis of land selling prices.

The comparable sales that are used shall be evidenced by an instrument recorded with the register of deeds of the county in which the land is located, if the date of such instrument and the recording date is not more than two years prior to the assessment year. [¶ 15.] Supplementing these statutes is SDCL 10-11-54 through 56, which specifically require that only arms-length sale transactions are to be included.

SDCL 10-11-56 defines an arms-length transaction as: [T]he term, arms-length transaction, means the transfer of property offered on the open market for a reasonable period of time between a willing seller and a willing buyer with no coercion or advantage taken by either party. The director of equalization shall analyze each sale to eliminate factors related to the sale which affect the sale price but which do not reflect the actual value of the real property. [¶ 16.] This Court has not had the opportunity to interpret SDCL 10-11-56.

However, a brief reference was made to the term, “arms-length transaction,” in the dissenting opinion of Richter Enterprises.

There, a transaction in which the purchasers bought a building on a contract for deed, after it had been offered for sale to the public at varying amounts for eighteen months, was treated as an arms-length transaction. 1997 SD 61, ¶ 29, 563 N.W.2d at 846 (Gilbertson, J. and Miller, C.J., dissenting). [¶ 17.] We agree with the Ohio Supreme Court’s holding that an arms-length transaction is characterized by three elements: it is voluntary, i.e., without compulsion or duress; it generally takes place in an open market; and the parties act in their own self-interest. Walters v.

Knox County Bd. of Revision, 47 Ohio St.3d 23, 546 N.E.2d 932, 935 (1989). Relying on Ohio’s definition, the Vermont Supreme Court held an intra-family transfer of property was not an arms-length transaction, because “[w]hile there is no doubt the sale was voluntary, it did not occur in an open market, and the price was negotiated by the parties without an appraisal.” Beach Properties, Inc. v. Town of Ferrisburg, 161 Vt. 368, 640 A.2d 50, 54 (1994). Moreover, there was insufficient evidence to show that self-interest prevailed over common family interests and predilections. Id.

We generally agree that an intra-family transfer of property is not an arms-length transaction. [¶ 18.] Here the record shows that County included intra-family sales of agricultural land in its market analysis. However, at the OHE hearing, County stated it only used these figures if an independent appraisal was first obtained. In order to learn whether an independent appraisal was conducted, County sent a verification letter to the parties, called them, or visited with them in person. County did not obtain a copy of the appraisal; it simply took the parties at their word that one had been conducted.

Nor did County require the parties to inform it of the appraised value, which could have been compared to the recorded sale price to verify that the sales price reflected true market value. [¶ 19.] This lax practice of verifying the sales price on intra-family transfers fails to meet the requirements for an arms-length transaction as defined at SDCL 10-11-56. The record contains the Certificates of Real Estate Value for nearly all transferred land in Perkins County in the 1998 assessment year.

The certificates contain information about whether there was a relationship between the parties and whether the property was offered for sale on the open market. Obviously that information is collected on each certificate in order to identify transactions which were not conducted at arms-length. Yet, despite these clear signals, County still chose to use four self-identified, intra-family, non-arms-length sales in its market analysis. The use of these transactions clearly violates the mandates of SDCL 10-11-56. [¶ 20.] That intra-family sales are used in the County’s market analysis is questionable.

As indicated, the problem is further compounded by the fact that County neither obtained a copy of the independent appraisal nor inquired of the appraised value. Because County did not obtain a copy of the independent appraisal when it chose to include intra-family sales in its market value, it violated the requirement that it “consider and document all elements of [a market analysis] that are applicable prior to a determination of true and full value.” SDCL 10-6-33. Townships have sufficiently overcome the presumption that County acted in accordance with the law.

OHE’s decision is supported by this record; the trial court’s reversal of that decision must be reversed. [¶ 21.] 2. County violated SDCL 10-6-33.6 by failing to properly re-evaluate the use of assessment zones. [¶ 22.] OHE made the following findings relevant to County’s failure to reevaluate the use of assessment zones for 1998: Finding # 12: The Director did not perform the analysis of determining (sic) whether the median value per acre in an identifiable region within the county deviates by more than ten percent from the county average for the 1998 assessment year.

Finding # 13: The Director did not determine if the deviation still existed to support the continued use of the four major zones in the county, or if more or less zones were needed. Finding # 16: Before the Director established the neighborhood zones for the 1998 tax year, he failed to perform the analysis to determine if first there was a ten percent deviation in any part of the county. [¶ 23.] At trial, the judge expressed concern with OHE’s findings that County did not perform a deviation analysis for 1998, because the record showed that County did indeed perform such a calculation.

Therefore, it ruled that OHE’s findings were clearly erroneous.

Although County did perform a deviation calculation for 1998, we reverse the trial court’s decision because the record also reveals that County incorrectly performed the deviation calculation. [¶ 24.] SDCL 10-6-33.6 provides: “If the median value per acre in an identifiable region within a county deviates by more than ten percent from the county average, the county director of equalization may establish a separate market value per acre for the land defined by the director of equalization within that region.” By its own admission, County understands this statute to mean that “if the median selling price in a certain area differs from the county average selling price,” the director of equalization is allowed to establish a separate assessment zone in that particular area. [¶ 25.] Our other opportunity to interpret SDCL 10-6-33.6 also originated from the creation of assessment zones in Perkins County.

In Lincoln Township, we held the so-called “zoning” or “neighbor-hooding” statute did not require the use of comparable sales to establish a separate top dollar value for each zone. 1996 SD 13, ¶ 13, 543 N.W.2d at 258. There, County increased the assessed value by 10 percent to establish the top dollar value for each zone, rather than using comparable sales values to arrive at such figure. [¶ 26.] Unlike the situation in Lincoln Township, where the question was whether a correct top dollar value had been established, here the issue is whether County correctly verified that the assessment zones were still valid.

Townships argue that in order to properly determine whether there has been a 10 percent deviation, County must compare the median selling price in a particular zone to the county average selling price. However, County does not do this. Instead, by its own admission, it compares the median assessment ratio every year.

Apparently, this has been County’s practice for some time, as revealed by a passage in Lincoln Township: The Board claims Seidel [the Director of Equalization] complied with the “neighborhooding” statute because the median sales value in zone one was $124.71 per acre and the median sales value of the entire county was $110.00 per acre. This was a deviation of more than 10 percent. The Board states: A simple application of the addition of 10 percent to the county median of $110.00 yields $121.00 per acre, the separate market value utilized by Sei-del. This is very, very close to the Zone 1 median of $124.71.

There are [ ] minor problems with this

interpretation. [ ] Seidel’s testimony does not indicate he determined the 10 percent assessment increase in that manner. Seidel testified he compared the sales ratios in each zone with the county sales ratio and discovered a difference of more than ten percent. The sales ratio in zone one was 62.7 percent (meaning land was assessed at 62.7% of its selling price), and the sales ratio for the county was 73.2 percent.

The statute provides the ten percent deviation must be in the “median value per acre.” It is immaterial here, however, because the median values per acre between zone one and the entire county did vary by 10 percent as required by the statute. In fact, the county median sales value was $110.00, while the zone one sales value was $124.71. Id. ¶¶ 9-10, 543 N.W.2d at 258. [¶ 27.] To further complicate matters, County presented a second explanation of how it conducted its deviation analysis.

The director of equalization stated at the OHE hearing that in order to satisfy the requirements of SDCL 10-6-33.6, he divided the assessed value of the land by the sales to assessment ratio, thereby establishing a median sales value per acre for each zone.

This figure, according to County, was then used to determine whether the assessment zones remained valid. [¶ 28.] Both the language of SDCL 10-6-33.6 and our decision in Lincoln Township make it clear: in order to properly determine whether the previously established assessment zones remain valid, County must annually compare the median sales value in a particular zone to the county average sales value. Notwithstanding this relatively clear statutory directive, the record shows that County compared the median assessment ratio every year, not the median sales value.

In short, County conducted a deviation analysis every year; it just did it incorrectly. [¶ 29.] Unfortunately, the record does not contain sufficient information to perform an accurate deviation calculation. As noted earlier, the sales data is skewed because it includes non-arms-length transactions.

Assuming all included sales records were valid, arms-length transactions, and that the sales data as given by County at the OHE hearing was correct, the following would be a deviation analysis for each zone: [¶ 30.] Thus, according to County’s own records and testimony, there is no longer a 10 percent deviation between Zone 1 and the county average, but there is now a 10 percent deviation between Zone 4 and the county average (Zone 4 is more than 10 percent lower than the county average). Zones 2 and 3 remain valid. [¶ 31.] These calculations show that County improperly maintained zones for the 1998 assessment.

Whether County used the incorrect formula to calculate the deviation, or whether it incorrectly calculated the figures using the correct formula, is immaterial. The point is that County’s 1998 assessment was incorrect because it was based on an improperly performed deviation analysis under SDCL 10-6-33.6. [¶ 32.] For the reasons set forth above, the trial court erred in. reversing OHE. However, OHE also committed error that must be addressed. As noted earlier, as part of its holding OHE directed that the top dollar value of Zones 2 and 4 be applied to Zones 1 and 3.

That is improper and fails to comply with the statutory directives for the assessment of property. It could result in arbitrary assessments, and lacks the uniformity and equality required by law. The appropriate remedy is to require County to perform the assessments in an appropriate manner, under the guidance set forth above. [¶ 33.] Therefore, the matter is reversed, and the trial court is directed to remand to County in accordance with this holding. [¶ 34.] SABERS, AMUNDSON, KONENKAMP, and GILBERTSON, Justices, concur. . The propriety of establishing assessment zones was upheld in Lincoln Township v. S.D.

Bd. of Equalization, 1996 SD 13, 543 N.W.2d 256. . A 1998 amendment to this

section added the words "location, size” preceding "soil.” Since November 1, 1997, was the valuation date for the 1998 assessment, we will consider the pre-amendment version of the statute. . Apparently, County considered whether a transfer fee was paid as a determinant of whether the transaction was conducted at arms-length.

At the OHE hearing, the Director of Equalization explained his rationale for choosing which sales records to submit to OHE: "I just told [the deputy] to copy anything that had a transfer fee or showed a transfer fee on it because if it doesn’t, why, it can’t be considered as an arms-length transaction anyway.” .. Assessment ratio equals the assessment value divided by the sales value. For example, a property that has an assessed value of $7,500 and a sales value of $10,000 would have an assessment ratio of .75. The "median assessment ratio” is the middle ratio of all the assessment ratios arranged in order. .

In Lincoln Township, the term "sales ratio” was used to denote the assessment-to-sales ratio, and here, the term used to describe the same ratio is "assessment ratio.” . We emphasize that the statute says, “county average,” not "county median.” There is a difference between an average value and a median value.