Montgomery Ward & Co., Inc. v. County of Hennepin

SIMONETT, Justice. This case involves valuation of a Montgomery Wards store for property tax purposes and returns to us after a second trial. We affirm the Tax Court’s valuation. *787 The property involved is the Montgomery Wards store in the Terrace Mall shopping center in Robbinsdale. The store is a traditional department store plus an automotive center, built in 1966, with a total of 181,111 square feet. Originally the store was freestanding; later a strip mall of 27 shops was added. It is undisputed that the Wards store has performed poorly because of poor location and other factors. As of January 2, 1987, 18 of the 27 shops were vacant. As of that date, the Wards property had an assessor’s estimated market value of $3,622,220. In this proceeding, Wards has challenged that valuation. At the first trial, the Tax Court found that the assessor’s estimated market value did not exceed the actual fair market value of the property ($3,695,700) but, after equalization, determined the adjusted market value on the books to be $3,289,178. At the second trial, resulting in the decision now before us, the Tax Court found the estimated market value to be $3,451,500, which, after equalization, resulted in an adjusted valuation of $3,071,835. In other words, the second trial resulted in the value of the property being reduced an additional $217,343. Wards, both at the first and second trial, has maintained that the estimated market value before equalization should be $2,900,000. In Wards’ appeal from the Tax Court’s first decision, we reversed and granted a new trial. Montgomery Ward & Co., Inc. v. County of Hennepin, 450 N.W.2d 299 (Minn.1990). We held that the Tax Court had committed reversible error by excluding certain evidence, and that Wards’ income approach “should have been given over-riding weight.” Id. at 308. On remand the Tax Court held two pre-hearing conferences to expedite discovery and discuss the issues to be considered. At the first trial, Wards had sought to obtain the appraisal and supporting data used in the Tax Court proceeding involving Donaldson’s department store in the Eden Prairie shopping center, claiming the information would support its own case. The County had objected to production of the information because of the Data Practices Act. Minn.Stat. §§ 13.50, 13.51 (1990). At the prehearing conference it was learned that Hennepin County had never had possession of this information, and therefore the requested information was not technically within the Data Practices Act. Hennepin County, however, obtained the information from Donaldsons’ attorney and turned it over to Wards. At the first trial, the Tax Court also sustained Wards’ objection to the County’s attempt to put in evidence Exhibits C, D, and E, which included the square foot rents paid by the small shops in Terrace Mall, because the County had failed to tell Wards that it intended to use these exhibits. At the prehearing conference, the Tax Court ruled that Exhibits C, D, and E would be received at the re-trial because surprise was no longer a factor. Finally, the Tax Court decided not to conduct an entirely new trial. Instead, the parties were allowed to supply any “missing links” in the record already made. Thus the table of rents from the appraisal in Eden Prairie Donaldsons and Exhibits C, D, and E were received, and the appraisal experts for Hennepin County and Wards were allowed to supplement their testimony. At the remand hearing, Gary Battuello (Wards’ expert) again used the percentage-of-income approach, supporting his use of 2.5 percent of retail sales with additional data from the appraisal in the Eden Prairie Donaldsons case. Kermon Benson (the County’s expert) again offered his allocation-of-income theory, resulting in a $3 per square foot value, and, relying on Exhibit C, pointed out that the rents for smaller stores in Terrace Mall ranged from $6 to $18.90 per square foot, with an average of $7.43 per square foot overall. In March 1991, the Tax Court issued its decision. The court found the market rent of Wards’ property to be $2 per square foot, rejecting Wards’ claim of $1.83 per square foot as too low. Then, adopting Mr. Battuello’s methodology (instead of Mr. Benson’s as it had in the first trial), the court computed market value as follows: Income ($2 X 181,111 sq. feet) = 362,222 less expenses: management (1%) ( 3,622) reserves ($10,000) (10,000) Net Operating Income = 348,600 Capitalized at 10.5% = 3,320,000 *788 The court then gave a weight of 75 percent to the income approach, 15 percent weight to the market approach, and 10 percent weight to the cost approach, and arrived at an adjusted market value after equalization of $3,071,835. Wards’ post-trial motions for amended findings or a new trial were denied. Wards now seeks appellate review here by writ of certiorari. Additional facts will be supplied as we go along. I. In Wards I, we ordered “a new trial wherein the court shall order discovery, admit into evidence and consider the requested data discussed in this opinion.” Id. at 308. Wards argues that this language limited the Tax Court’s authority to admit only the Eden Prairie Donaldsons appraisal data, and it was without authority to admit other data, specifically Exhibits C, D, and E, which had been excluded at the first trial. There is no merit to Wards’ argument. The effect of a remand order depends on the facts of each case. Klampe v. Klampe, 145 Minn. 404, 405, 177 N.W. 629, 630 (1920). In this case, we placed no restrictions on the remand, but did say that the Eden Prairie Donaldsons appraisal (the contents of which were then unknown) should be received. The Tax Court, trying this case as a bench trial, saw no need to re-do what had already been done, but chose to use the record already made, supplementing it with whatever other evidence was relevant. This approach, it seems to us, was sensible and within the Tax Court’s discretion. See Part 11(B), infra. Exhibits C, D, and E had been excluded at the first trial because of unfair surprise to Wards. That objection no longer existed at the time of the second trial, and the exhibits were properly received. II. We now reach the main issue: Does the evidence sustain the Tax Court’s findings of fact and conclusions of law? This court will not disturb the Tax Court’s findings on appeal where sufficient evidence supports the decision. Nagaraja v. Commissioner of Revenue, 352 N.W.2d 373, 376 (Minn.1984). Even if there is evidence to support factual findings, this court may, however, order a reversal if, upon reviewing the entire evidence, it is left with a firm conviction that a mistake has been made. City of Minnetonka v. Carlson, 298 N.W.2d 763, 766 (Minn.1980). At the second trial, pursuant to our mandate, primary emphasis was placed on the income approach to property valuation. This approach is based on the principle of anticipation, that a buyer of an income-producing asset will pay an amount equal to the income that the property should reasonably be expected to generate, minus expenses, divided by a capitalization rate that investors would reasonably expect to obtain. Northerly Centre Corp. v. County of Ramsey, 248 N.W.2d 923, 925 n. 2 (Minn.1976). In the case of retail stores, this income is typically rental income. With property like the Wards store, the difficulty lies in the fact that the property is owner-occupied, so that the appraisal must impute a constructive rent. In arriving at a constructive rent, Mr. Battuello and Mr. Benson not only arrived at different figures but used different methodologies. Mr. Benson used an “allocation-of-income” methodology. Under this theory, the appraiser examines “comparable” rents of anchor tenants in other malls and of freestanding stores (rent per square foot). In addition, it is assumed that anchor tenants in malls pay a disproportionately low amount of rent, because smaller mall tenants typically pay the landlord a surcharge in order to be near a well-known anchor. Mr. Battuello, on the other hand, used a “pereentage-of-income” approach. This theory considers that the rent for a department store approximates a certain percentage of the retail sales generated by the store. The annual retail sales figure for past years is a known factor. Typically, the tenant and landlord then negotiate a rental percentage figure, say, 2.5 percent of sales (to use the figure urged by Wards in this case). Rent is then paid on a two-step basis. First, the tenant pays 2.5 percent of a specified figure for annual retail sales, say 2.5 percent of $13,250,000, to use *789 the annual retail sales level at the Wards store for 1987. This produces a rent of $331,250, or $1.83 a square foot. This is called the base rent. It is a fixed rent, i.e., if the sales level falls below $13,250,000, the tenant still pays $331,250 in rent. Next, if the sales level in any year of the lease exceeds $13,250,000 the tenant pays an additional rent of 2.5 percent on the sales which exceed the “breakpoint” of $13,250,000. (Ordinarily the stabilized annual retail sales level is considered to be a “natural” breakpoint.) This additional rent is called overage rent. The same percentage figure of 2.5 percent, it should be noted, applies to both base rent and overage rent. If, however, the tenant must pay 2.5 percent of $13,250,000 in a year in which the retail sales fall short of that figure, the tenant actually pays in rent a percentage higher than 2.5 percent. For the second trial, the Tax Court entirely adopted the percentage-of-income approach used by Wards’ expert but used different figures. In this appeal, Wards challenges the Tax Court’s decision on three grounds. First, Wards contends that the Tax Court erred in substituting a percentage of 2.7 percent for the 2.5 percent adopted by Mr. Battuello. Second, Wards claims the Tax Court failed to make certain deductions for vacancy and expenses in arriving at a net rental income. Finally, Wards claims the Tax Court failed to give “over-riding weight” to the income approach as mandated by this court. The Tax Court set out its reasoning in two memoranda, the first attached to its “amended” order for judgment and the second to its order denying Wards’ post-trial motions. The parties have treated these memos as containing the court’s findings and conclusions, as do we. We now turn to Wards’ specific challenges. A. Wards says the Tax Court should have used Mr. Battuello’s 2.5 percent figure. Mr. Battuello supported his figure by citing a publication entitled Dollars and Cents, which showed that for a national chain anchor in a regional shopping center, the percentage figure ranged from 1 percent to 3 percent. In addition, Wards noted that Judge Roemer had used a percentage of 2.5 percent in the Eden Prairie Donaldsons case. Finally, Mr. Battuello listed 10 department stores in the Midwest where the percentage varied from 1.5 percent to 2.5 percent. He obtained some of his data for the 10 listed department stores from the appraisal report in the Donald-sons case, which gave lease data on 13 department stores also located in the Midwest. There are weaknesses, however, in this supporting data. First of all, as the Tax Court noted, the Wards store at Terrace Mall does not fit very well within the Dollars and Cents category of a regional shopping center. The Wards store, with an automotive center, is the only anchor tenant in a mall with fewer smaller stores than one would expect in a regional mall. The Donaldsons store, by way of contrast, was one of three anchor tenants at Eden Prairie (the other two being Sears and Target) in a mall with space for 125 smaller stores. As for the 10 “comparable” department stores listed by Wards’ appraiser, the data is not complete and in some instances can be manipulated. For example, for the J.C. Penney store in Marshfield, Wisconsin, the percentage rent is listed as 1.5 percent and the base rent per square foot at $3.10 (a base rent of $153,184 based on $3.10 times 52,640 square feet). It turns out, however, that total retail sales from this store were $3,316,320, which works out to $63 a square foot, or a percentage of retail sales paid for base rent of 4.9 percent. Apparently, the store’s sales did not reach break-point, and, as we have noted, the percentage rent figure is only reliable in indicating base rent as a percentage of sales if break-point is reached. There were other problems with the data. For two K-Mart stores, no percentage figure was listed. Moreover, Mr. Battuello relied on the Don-aldsons appraisal for his report in this case and, in that appraisal, the level of retail sales and the information on base rent per square foot and percentage rent was from different years. This disparity caused Judge Roemer in the Donaldsons case to characterize the comparables used in the *790 Donaldsons appraisal as “seriously flawed.” It appears that multiplying 2.5 percent times the retail sales level may lead to a false base rent per square foot (or per square foot value) for the Wards property. Apparently this possibility troubled Judge Gustafson in this case. The judge noted that the average per square foot value (or base rent per square foot) of Mr. Battuel-lo’s comparable stores was $2.29, compared with the $1.83 figure proposed by Wards. The Dollars and Cents survey showed the median base rent per square foot for national chains in regional shopping malls to be $2.83. These figures suggested to the Tax Court that $1.83 per square foot was low and that the percentage rent figure should be increased. The Tax Court, therefore, increased the per square foot value of the property to $2.00 a square foot (or the same in base rent), which works out to 2.7 percent of retail sales instead of Wards’ estimated 2.5 percent. We conclude that the Tax Court’s use of $2 per square foot (translating to 2.7 percent) is supported by the evidence and should not be disturbed. The Tax Court also noted that the smaller tenants in Terrace Mall paid an average rent of $7.43 a square foot, and Wards points out these rentals lack any probative value, at least under the percentage-of-income methodology that was used. The Tax Court agreed, however, that these rentals do not determine the rate for large department stores and said they were useful only to give some idea of the general range for leased retail space in the area. B. In the first trial, the Tax Court, in using the income approach, applied the “allocation of income” methodology employed by Hennepin County’s appraiser. Mr. Benson had deducted 10 percent for vacancy and 10 percent for management and reserve expenses. The Tax Court used these figures but raised the vacancy rate to 15 percent. In the second trial, after arriving at gross rental income under the percentage-of-income approach, the Tax Court made no deduction for vacancy, deducted only 1 percent for management expenses, and a flat $10,000 for reserve expenses. Wards assigns error. The trouble with Wards’ position is that its own expert had testified, at the first trial, to no vacancy rate, only 1 percent for management expense, and a flat $10,000 deduction for reserves. The Tax Court simply adopted the testimony of Wards’ own expert. Wards seems to argue that the Tax Court should be bound by its earlier findings on vacancy and expense deductions, absent additional evidence on these issues at the trial on remand. In other words, Wards wants the best parts of both Mr. Benson’s and Mr. Battuello’s testimony. We see no reason why this must be so. Because a new trial was ordered, the case occupies the status it had before the first trial. Muggenburg v. Leighton, 240 Minn. 21, 25, 60 N.W.2d 9, 11 (1953); Wilson v. Anderson, 145 Minn. 274, 276, 177 N.W. 130, 131 (1920). Thus, upon reversal, the original findings were no more, at least not until re-adopted. The Tax Court decided to hold a limited hearing on remand to supplement the record already made, rather than rehear all of the same evidence as well as new evidence. This seems appropriate enough in a bench trial such as this one. See Nash v. Kirschoff, 161 Minn. 409, 411, 201 N.W. 617, 618 (1925); cf. Minn. R.Civ.P. 59.01(g). The question before us is not whether there was sufficient new evidence submitted at remand to support the Tax Court’s change of position, but whether the record as a whole supports the court’s findings. We think it does. Both at the first trial and again at remand, Mr. Battuello testified he did not use a vacancy factor because Wards completely occupied the property. This evidence is sufficient to support the Tax Court’s omission of any vacancy factor. As to the management and reserve figures, Mr. Benson did not offer any explanation for the figures he used at the first trial, and, on remand, he was not asked for any explanation. At the first trial, Mr. Battuel-lo estimated a 1 percent management fee and explained in some detail his allowance of a $10,000 reserve (for a one-time roof replacement that would be needed). It *791 seems the Tax Court, on remand, simply adopted Mr. Battuello’s methodology and, as to deductions, his numbers. We cannot say this was error. III. We instructed the Tax Court on remand to give “over-riding weight” to the income approach in light of the serious weaknesses in the cost approach (at best yields a ceiling only) and the market approach (no truly comparable sales). Wards contends the Tax Court violated this instruction, and, in any event, the evidence is insufficient to support the allocation of weight assigned by the Tax Court to the^ three approaches. We find this argument unpersuasive. The Tax Court allocated 75 percent weight to the income approach, 15 percent to market, and 10 percent to the cost approach. We think this complies with our instructions. As the Tax Court noted, it gave the income approach three times the weight of the other two approaches combined. Moreover, we think the record supports the Tax Court’s weight allocation. If the cost and market approaches had inherent weaknesses in arriving at a fair market value in this case, the income approach, while the best measure of value, was not without its own weaknesses. Because many large retail stores are owner-occupied, an accurate estimate of rental income is difficult to obtain. Nothing in our remand required the Tax Court to consider the income approach to the exclusion of the other two approaches. Cf. Northerly Centre Corp. v. County of Ramsey, 248 N.W.2d 923, 927 (Minn.1976) (“[I]t is not mandatory upon the trier of fact to accept any particular valuation approach as the sole basis for determining market value,” and consequently the tax judge was not limited to the income approach). In this case, the experts for both sides used all three approaches, although on remand focusing on the income approach. It was for the Tax Court to consider the entire record and, in doing so, it was not error to give less than 100 percent weight to the income approach. Finally, the tax judge’s use of all three approaches in his final calculation of market value is consistent with appraisal procedure and case law. The three approaches are supposed to act as a check against one another. Am. Inst, of Real Estate Appraisers, The Appraisal of Real Estate 561-66 (1987). As this case illustrates, real estate appraisal is at best an imprecise art, and a tax court proceeding is not high-low arbitration where the decision-maker must choose the figure submitted by one or the other party. The Tax Court brings its own expertise and judgment to the hearing, and its valuation need not be the same as that of any particular expert as long as it is within permissible limits and has meaningful and adequate evidentiary support. Northerly Centre Corp. v. County of Ramsey, 248 N.W.2d at 927. Affirmed.