Opinion
DA 23-0030
09-24-2024
WILLIAM M. SOLEM and ELLEN G. SOLEM, Plaintiffs, Appellees, and Cross Appellants, v. STATE OF MONTANA, DEPARTMENT OF REVENUE, Defendant, Appellant, and Cross Appellee.
For Appellant: Nicholas J. Gochis, Senior Tax Counsel, Montana Department of Revenue, Helena, Montana Stefan T. Wall, Wall, McLean &Gallagher, PLLC, Helena, Montana For Appellees: Lon J. Dale, Rachel H. Parkin, Milodragovich, Dale &Steinbrenner, P.C., Missoula, Montana Dylan McFarland, Knight Nicastro Mackay, LLC, Missoula, Montana
Submitted on Briefs: September 27, 2023
APPEAL FROM: District Court of the Eleventh Judicial District, In and For the County of Flathead, Cause No. DV-10-073 (D) Honorable Dan Wilson, Presiding Judge
For Appellant: Nicholas J. Gochis, Senior Tax Counsel, Montana Department of Revenue, Helena, Montana Stefan T. Wall, Wall, McLean &Gallagher, PLLC, Helena, Montana
For Appellees: Lon J. Dale, Rachel H. Parkin, Milodragovich, Dale &Steinbrenner, P.C., Missoula, Montana Dylan McFarland, Knight Nicastro Mackay, LLC, Missoula, Montana
JAMES JEREMIAH SHEA JUDGE
¶l The State of Montana, through the Department of Revenue ("DOR" or "Department"), appeals from the Eleventh Judicial District Court, Flathead County's October 15, 2019 Findings of Fact, Conclusions of Law and Order finding in favor of Plaintiffs, William and Ellen Solem ("Solems"). The Solems cross-appeal the District Court's May 7, 2021 Order denying their Motion to Amend the Class Definition.
¶2 We address:
Issue One: Whether the District Court erroneously substituted its opinion for that of the Department of Revenue's when it ruled that the Department of Revenue's appraisal methodology was unlawful.
Issue Two: Whether § 15-l-407(2)(a), MCA's payment-under-protest requirement is unconstitutional.
¶3 We reverse and remand for proceedings consistent with this Opinion.
FACTUAL AND PROCEDURAL BACKGROUND
¶4 This dispute arises from DOR's 2008 mass appraisal of Flathead County's "Neighborhood 800" lakefront properties. The Solems own property in Neighborhood 800. In both 2002 and 2008, DOR used a mass appraisal approach for valuing Neighborhood 800 properties, which allows it to value multiple properties at once-in contrast to a fee appraisal approach, which values properties one at a time.
¶5 The mass appraisal process generally progresses through the following steps:
(1) An appraiser considers four factor types-(1) physical; (2) economic; (3) governmental; and (4) social-to group properties into a homogeneous "neighborhood."
(2) Appraisers visit the neighborhood in-person to "verify" recent land sales by talking to sources such as buyers, sellers, realtors, and local appraisers. This verification includes both the notation of the characteristics that influenced the purchase price of the property, such as topography, location, and surroundings, as well as a determination of whether the sale was a "market," or "arms-length," transaction. Only market transactions are used later in the process.
(3) An appraiser creates a hypothetical property, denoted a "base lot," using the data collected during the verification process and a computerized "sales comparison method." The base lot is intended to be a representative average of the neighborhood in terms of price. The appraiser uses his or her judgment in selecting which sales go into the model to exclude any "outliers" that would make the base lot unrepresentative, for instance if the sale was for significantly higher than market price.
(4) The appraiser builds a model around the base lot values and tests that model by comparing it to its verified sales data to determine whether the model would accurately predict the sales prices of the recently sold properties. Using his or her judgment, the appraiser quality checks the model by introducing and removing variables from the model until its predictions reach an acceptable accuracy level.
(5) DOR applies the model to each property in the neighborhood to determine its appraised valuation.
(6) DOR compares the unique "influence factors" of each property, such as topography, restricted view, or access problems, and staff appraisers adjust each individual valuation based on their judgment and market data.
(7) DOR adjusts the valuation again if a property owner informally challenges the valuation pursuant § 15-7-102(3), MCA, otherwise known as an AB-26 protest ("AB-26").
¶6 DOR's 2002 appraisal of the Solems' land was $229,500. Real estate prices increased substantially over the next several years, and by 2008 Flathead Lake properties were in the midst of a buying frenzy. DOR's 2008 appraisal of the Solems' property was $1,233,050. DOR received 187 informal appeals from the 2008 Neighborhood 800 reappraisal, with five of those appeals reaching the Montana Tax Appeal Board ("MT AB"). MTAB ruled the Department's methodology and valuations were correct. Instead of appealing DOR's determination to the MTAB, the Solems sued DOR in 2010, seeking approximately $450 they alleged they had overpaid in taxes.
¶7 In 2016, the District Court certified the Solems' case as a class action. The class included Neighborhood 800 lakefront property owners who had paid their taxes under protest. The District Court later bifurcated the case into separate liability and damages phases. DOR's appeal concerns only the District Court's liability determination.
¶8 The District Court held a four-day bench trial on liability issues from March 11-14, 2019. At trial, the Solems argued that "the DOR's method of assessing lakefront property [was] improper and unlawful" because it only took into account "three variables; the price per front foot, the depth of the lot, and a qualitative variable." DOR argued that its "sales comparison" methodology was lawful, and variable selection was merely a part of the "process" of using the methodology to produce appraisals.
¶9 DOR called Regional Manager Scott Williams, who oversaw the appraisal process for Neighborhood 800, to explain the judgment-based decisions he made at each step of the mass appraisal process. Williams testified that at step three he decided to remove 17 outlier sales from the base-lot data pool because leaving them in would have "super-valued the Solems' lot, and everybody else's lot." At step four, he testified that he selected the three variables he used-price per front foot, lot depth, and a qualitative binary variable- because they produced a model that was 83% accurate and which could not be improved by adding or subtracting other variables. He explained that he opted not to include a negative price adjustment in the qualitative variable at step four because that made the model less accurate, but that lots with negative characteristics were readjusted downward at step six.
¶10 To support their argument, the Solems called three licensed appraisers as expert witnesses: Steven Hall, Cynthia Smith-Paige, and David Lennhoff.
¶11 Hall testified that "the appraisal profession has three valuation methodologies: the cost approach, the income capitalization approach, and the sales comparison approach." He testified that the sales comparison methodology was one he often used in evaluating lake-front property along Flathead Lake. He opined that DOR's use of the "front footage" variable in its model was "absolutely correct," but he would have adjusted for size differently using a less linear regression. He stated that he would not have used the "lot depth" variable DOR chose but would have used several other variables in its place to approximate the land's fitness as "a building site." He admitted that he did not understand DOR's third variable. Hall explained his own appraisal approach that resulted in a $1,020,000 valuation of the Solems' property.
¶12 Smith-Paige testified that DOR correctly selected the mass appraisal approach, the sales comparison method, a base lot model, and "the unit of measure that they used for the front foot." She agreed that the Department's decision to use a qualitative variable was appropriate, but that the decision not to include a negative value in the variable reduced the model's "credibility." She asserted that the fifth step in the process, adjusting for unique influence characteristics, would not correct for this reduction in credibility. She testified that the 29-property sample size was too small, and that DOR should have worked to expand the size by including some of the outlier properties it had excluded or including properties outside of the neighborhood. She opined that DOR should have reassessed its entire mass appraisal when approximately 30% of the property owners in Neighborhood 800 filed AB-26s. Smith-Paige admitted that 2008 was "a difficult market for everyone" to value.
¶13 Lennhoff testified that it was appropriate to use the mass appraisal approach, the sales comparison method, and the base lot model, and it was "mandatory" to quality check the model by testing comparator sales. He asserted that it was inappropriate to use a binary variable as a stand in for qualitative sales characteristics but acknowledged that he was "not a model builder" and did not know which other variables he would have added. He agreed that DOR's construction of its base lot was "fine," including the use of lot depth as a variable. He stated that the number of appeals filed after the 2008 appraisal was not a definitive indicator that there was a problem with the methodology. Summing up his testimony, he stated, "[T]here are a lot of things about [the process] that I think are fine. If s just that the way, ultimately, it's applied here simply does not work."
¶14 The District Court held that DOR's mass appraisal methodology was unfair and unconstitutional, finding that although the base-lot model used for DOR's mass appraisal methodology is generally accepted, the variables DOR used were not. The District Court faulted the Department for removing 17 "outlier" verified sales, believing that instead of eliminating these outliers DOR "should have proposed additional variables into the model to account" for these sales. It awarded damages, costs, and fees to plaintiffs. Following the District Court's October 2019 order, the parties agreed to a stipulated final judgement which reserved their right to appeal.
STANDARDS OF REVIEW
¶15 We review a district court's findings of fact for clear error. State v. Covington, 2012 MT 31, ¶ 13, 364 Mont. 118, 272 P.3d 43 (citing State v. Pearson, 2011 MT 55, ¶ 11, 359 Mont. 427, 251 P.3d 152). We review a district court's legal conclusions for correctness. Covington, ¶ 13.
DISCUSSION
¶16 Issue One: Whether the District Court erroneously substituted its opinion for that of the Department of Revenue's when it ruled that the Department of Revenue's appraisal methodology was unlawful.
¶17 We have cautioned that "it is not a judicial function to act as an authority on taxation matters." Dep't of Revenue v. Grouse Mountain Dev., 218 Mont. 353, 355, 707 P.2d 1113, 1115 (1985). We have long held that DOR appraisals are presumed correct with the burden resting on taxpayers to disprove them. W. Airlines v. Michunovich, 149 Mont. 347, 353, 428 P.2d 3, 7 (1967). A district court may not engage in a "wholesale substitution" of its opinion for the opinion of the agency. Peretti v. Dep't of Revenue, 2016 MT 105, ¶ 17, 383 Mont. 340, 372 P.3d 447 (citing O'Neill v. Dep't of Revenue, 2002 MT 130, ¶ 23, 310 Mont. 148, 49 P.3d 43).
¶18 Section 15-7-103(1), MCA, requires DOR to provide for "a general and uniform method of appraising" property, and § 15-8-111(1), MCA, requires the method to appraise property "at 100% of its market value." However, we have held that DOR's "use of various approaches to valuation does not contravene the mandate that appraisals be done by the same 'method' statewide." CHS, Inc. v. Dep't of Revenue, 2013 MT 100, ¶ 20, 369 Mont. 505, 299 P.3d 813. This includes DOR's employment of a "market-based method utilizing a combination of approaches." CHS, ¶ 20. We have also recognized that any "method of assessing property and estimating market values is by no means perfect" and perfect 100% valuation is "unattainable." Albright v. State, 281 Mont. 196, 213, 933 P.2d 815, 826 (1997).
¶19 DOR argues that when the District Court ruled that the Department's appraisal methodology was unlawful, it improperly substituted its opinion for that of DOR. The Department points to our recent decision in Peretti as instructive. In Peretti, landowners who also owned property on Flathead Lake disputed DOR's 2012 appraisal value of their residential lots. Peretti, ¶¶ 3-4. MTAB upheld the appraisals on appeal. It held that the Perettis' comparison of DOR's mass-appraisal methodology and sales data from allegedly comparable properties was insufficient to render the appraisal incorrect. Peretti, ¶¶ 10-12. The Perettis appealed MTAB's decision to the district court, which found that MTAB committed clear error and ordered MTAB to accept the Perettis' valuation and return taxes paid under protest. Peretti, ¶ 13. We reversed and held that the Perettis had failed to overcome the substantial burden to disprove the accuracy of DOR's appraisal. Peretti, ¶19.
¶20 The Solems argue Peretti is distinguishable from this case because the Perettis challenged DOR's appraisal based on individually flawed assessments, not defects in DOR's overall methodology. The Solems' efforts to distinguish Peretti are unavailing. In Peretti, we explicitly addressed the district court's finding that "the methodology of the DOR" was flawed and held that because "DOR's appraisal was based upon a proven computer model that incorporated data on the verified sales of 29 other properties to reach the appraised valuation," it was "presumed to be correct and the Perettis bore a substantial burden to disprove it." Peretti, ¶¶ 12, 19.
¶21 The methodology at issue here is based on the same 29-sale model at issue in Peretti and is entitled to the same presumption of correctness. As in Perretti, the evidence the Solems offered at trial to rebut the presumption was insufficient to meet their burden. Each of the Solems' experts testified that the mass appraisal approach, the sales comparison method, and the base lot model DOR employed were appropriate, widely accepted methodologies.
¶22 There was notable disagreement between the Solems' experts as to the alleged flaws in DOR's methodology. While Smith-Paige stated that the high rate of AB-26 filings meant DOR should have entirely reworked its model, Lennhoff testified that the number of filings did not necessarily indicate any flaws. Hall testified that lot depth was too imprecise a measure for building potential, but Lennhoff took no issue with DOR's use of lot depth in its valuation. Smith-Paige and Lenhoff both identified the qualitative variable as a source of concern, but while Smith-Paige would have only added one more dimension to it, Lenhoff was unsure what to swap it out for. Ultimately, the only point of agreement the Solems' experts had was their opinion that DOR should have exercised its judgment differently when accounting for different variables and including or excluding sales from its model. All this testimony goes to show is what we have recognized for decades: appraisal in this fashion is a difficult undertaking, which is why DOR's task is limited to "a reasonable attempt to equalize appraisal of real property throughout the State" that "comports with the most modem and accurate appraisal practices available." Albright, 281 Mont, at 213, 933 P.2d at 826.
¶23 To cut through the confusion of competing recommendations from the Solems' experts, the District Court homed in on one metric: the R squared value. DOR's expert appraiser, Richard Hagar, testified that the R squared value indicates "confidence level," meaning the higher the R squared value, the more confident the DOR is in its model. The District Court pointed out that when the outlier properties Williams had excluded from the analysis at step five were reinserted into the data, the R squared of the model plummeted. However, Williams testified that the inclusion of these "outliers" would have "super-valued" the Solems' property, leading to an even higher tax burden. Williams also testified that the R squared value is only one of several ways that DOR validated its model; another was comparison between the model's output and other appraisals. One such comparison is available in the record here. At trial, William Solem testified that he secured a loan on his property based on a November 2008 evaluation by an appraiser named Steve Backer. Backer's appraisal, completed four months after DOR's, valued the property at $1,250,000-$16,950 more than DOR's appraisal. The District Court's selection of one accuracy metric over another when none of the experts who testified before it did so is precisely the sort of inexpert judgment substitution we have cautioned against in the past. See Grouse Mountain Dev., 218 Mont, at 355, 707 P.2d at 1115; Peretti, ¶ 17.
None of the Solems' experts relied on the R squared value as an indicator of the alleged inaccuracy of DOR's model. Nor did Williams base his confidence level in the model on it. He merely brought it up when he explained how DOR continuously checks the accuracy of the model as it is being constructed.
¶24 The District Court erred by substituting its judgment for that of DOR. While other appraisers may have made different decisions regarding particular variables included in DOR's methodology, the record establishes that DOR employed a consistent, accepted process for arriving at market value. Having determined DOR's methodology was adequate and not arbitrary, we also conclude the District Court erred in concluding DOR's methodology ran afoul of the Constitution.
¶25 Issue Two: Whether § 15-1-407(2)(a), MCA's payment-under-protest requirement is unconstitutional.
¶26 The Solems argue that § 15-1 -407(2)(a), MCA's requirement that all class members "must have paid the [challenged] tax under protest" is unconstitutional. We have long held that the courts should avoid constitutional issues whenever possible. Ingraham v. Champion Int'l, 243 Mont. 42, 46, 793 P.2d 769, 771 (1990) (citing State ex rel. Hammond v. Hager, 160 Mont. 391, 400, 503 P.2d 52, 57 (1972)). This is particularly true where the constitutionality of a legislative act is at issue. Ingraham, 243 Mont, at 46-47, P.2d at 771 (citing Shea v. North-Butte Mining Co., 55 Mont. 522, 530, 179 P. 499, 501 (1919)). The basis for a class action in this case was the allegation that DOR employed a methodology that unlawfully valued multiple properties in Neighborhood 800. Having concluded that DOR's methodology was not unlawful, we need not consider the Solems' constitutional challenge to § 15-l-407(2)(a), MCA.
CONCLUSION
¶27 The District Court erred by substituting its judgment for that of DOR in determining that DOR's model failed to value properties lawfully. The District Court's October 15, 2019 Order is reversed.
We Concur MIKE McGRATH LAURIE MCKINNON INGRID GUSTAFSON DIRK M. SANDEFUR