OPINION OF THE COURT BY KLEIN, J.

The City and County of Honolulu (the City) appeals from a Tax Appeal Court ruling on the assessed fair market value of the ocean-front Black Point Road property of Mary P. Steiner, trustee (Steiner) for the 1988-89 and 1989-90 tax years. The City objects to the valuation of the Tax Appeal Court because: (1) the City believes that the court disregarded sales to Japanese buyers in determining fair market value, and (2) the court applied the Kahala Beach benchmark for a 20,000 sq. ft. lot to only 14,000 sq. ft. of the Steiner property without adjusting the benchmark upward to reflect the decreased size. Steiner cross-appeals from the Tax Appeal Court's finding that 14,000 sq. ft. of the property is usable, claiming that only 11,000 sq. ft. of the 46,707 sq. ft. lot is usable. This appeal concerns the valuation of the land only, not the improvements thereon.

Our previous decisions specify the applicable standard of review:

It is well settled that in reviewing the decision and findings of the Tax Appeal Court, a presumption arises favoring its actions which should not be overturned without good and sufficient reason. The appellant has the burden of showing that the decision of the Tax Appeal Court was "clearly erroneous."

In re Puna Sugar Co., 56 Haw. 621, 623, 547 P.2d 2, 4 (1976) (citing In re Ewa Plantation Co., 47 Haw. 41, 51, 384 P.2d 287, 292 (1963)); see also In re Hawaiian Land Co., 53 Haw. 45, 49, 487 P.2d 1070, 1074 (1971), appeal dismissed, 405 U.S. 907 (1972); In re O.W. Ltd. Partnership, 4 Haw. App. 487, 492, 668 P.2d 56, 61 (1983).

I. A. Real Property Tax Valuation

Under the Revised Ordinances of Honolulu 1990 (ROH) chapter 8, article 7), the Director of Finance of the City and County of Honolulu is responsible for the valuation and assessment of property for the purposes of real property taxation. ROH § 8-7.1(a) provides that:

The Director of Finance shall cause the fair market value of all taxable real property to be determined and annually assessed by the market data and cost approaches to value using appropriate systematic methods suitable for mass valuation of properties for taxation purposes, so selected and applied to obtain, as far as possible, uniform and equalized assessments throughout the county. . . .

We have stated the general principles of tax valuation as follows:

The `ultimate purpose of valuation, whether in eminent domain or tax . . . proceedings is to arrive at a fair and realistic value of the property involved.' Our statutory scheme of real property taxation adopts this principle by identifying `fair market value' as the relevant measure of the value of property. And we have previously defined `market value' as `the value in money of any property for which that property would sell on the open market by a willing seller to a willing buyer.'

In re Amfac, Inc., 65 Haw. 499, 502, 654 P.2d 363, 366 (1982) (citations omitted). We review the valuation of the Steiner property in light of these principles.

The Department of Finance has established more specific guidelines for real property valuation in the Procedure and Reference Manual of the Real Property Assessment Division, Department of Finance, City and County of Honolulu (Procedure and Reference Manual or Manual). The Manual provides two basic methods of land valuation, the market data approach and the income approach (used for income generating properties). Procedure and Reference Manual §§ 991.00 — 912.00. The market data approach is described as follows:

[T]his method involves obtaining all available sales data, qualifying these as being transacted at `arms-length' conditions or being `bona fide' transactions, and comparing these recently sold properties to the one being appraised, making adjustments as necessary for comparisons. Unit prices demonstrated by actual sales, either adjusted or unadjusted, provide valid estimates of market value. Adjustments for dissimilarities in the market data approach to value are made by plus or minus of dollar amount or percentages, from the comparable parcels to the subject, benchmark or typical parcel in the neighborhood. The major factors which should be considered in the adjustments include the time of sale, terms of sale, location of compared parcels, differences in physical characteristics among them, and all other factors which might have been significant so as to influence the prices paid for the properties.

Procedure and Reference Manual § 911.00.

Section 941.31 of the Manual directs assessors to consider three general factors affecting the value of urban and suburban residential lands: location, neighborhood characteristics and site characteristics. Site characteristics include: (1) size and shape of lot; (2) topography and soil conditions; (3) landscaping; and (4) accessibility. Another section of the Manual notes that the "topography may be such that expensive excavations are required." Procedure and Reference Manual § 913.00. Under these circumstances the appraiser is instructed to note any adjustments for these conditions on the appraisal card as a percentage reduction in the value of the affected parcel. Id.

In assessing shoreline properties, the "most important factor of value" is the quality of the shoreline itself. Procedure and Reference Manual § 941.34. Appraisers are advised to consider "the amount, kind, and quality of shoreline, the quality and condition of the surf and water, the view, and the accessibility to the shore, and including the hazards or detriments because of the natural wave action." Id.

B. The City's Valuation

The primary method used by the City tax appraiser to value the Steiner property was to establish (1) a "typical" lot size and (2) a "benchmark" value, in dollars per square foot, for the particular neighborhood or area. The benchmark value was calculated by reviewing the sales data for ocean-front property in the surrounding areas from the year before last (e.g. 1986 sales data for the 1988 tax year). If, as in the case of the Steiner property, a lot were larger than the typical lot, the excess area or "overage" would be valued per square foot at a percentage of benchmark value. If a lot were smaller than the typical lot, the benchmark would be adjusted upward to reflect the greater value per square foot.

For the 1988 and 1989 tax years, the City assessed the property based on a typical lot size of 20,000 sq. ft. The assessor, Irene Nakamura, applied the Black Point benchmark to 20,000 sq. ft. and 50% of benchmark to the remaining 26,707 sq. ft. The benchmark value, based on comparable sales data, was $75 per sq. ft. for 1988 and $95 per sq. ft. for 1989 in the Black Point area. Based on this formula, she assessed the property as follows:

1988

100% x $75 x 20,000 sq. ft. = $1,500,000 50% x $75 x 26,707 sq. ft. = 1,001,513 __________ $2,501,513

100% x $95 x 20,000 sq. ft. = $1,900,000 50% x $95 x 26,707 sq. ft. = 1,268,583 __________ $3,168,583

For the 1988 benchmark, the City relied heavily on the 1986 sales of three Black Point properties which it considered comparable to the Steiner property. These properties were calculated to have land values of $83.66, $110.27 and $82.56 per sq. ft. The City also considered sales of ocean-front properties in the Diamond Head and Kahala areas that ranged from $81.08 to $155.53 per sq. ft. From these sales data, the City determined that the benchmark should be $75 per sq. ft. for the Black Point area. Ms. Nakamura testified that the Black Point benchmark also reflected a discount for the topography in the Black Point area; she made no other adjustments to the valuation of the Steiner property based on its topography.

Similarly, the 1989 Black Point benchmark of $95 per sq. ft. was based on 1987 sales in the surrounding areas, including the sale of one Black Point property.

The City used the same 50% (of benchmark) overage factor for all ocean-front lots which exceeded 20,000 sq. ft. in both the Black Point and Kahala Beach areas. No adjustment was made based on the topography or usability of the overage area. Thus, level (buildable) beach-front overage on Kahala Avenue was valued at only $10 per sq. ft. more than steep, boulder strewn (unbuildable) overage with no beach access such as on the Steiner parcel. Both the City and Steiner argued about whether there was evidence that certain lots were level and 100% usable and whether these lots were given a 25% overage adjustment in past years. However, a careful review of the record only reveals evidence that:

(1) the Steiner property and the adjoining two parcels had large areas of rocky unusable land;

(2) these three parcels were previously subject to a three part valuation, 100% of benchmark for 10,000 sq. ft., 50% for 1,000 to 1,780 sq. ft., and 25% for the remaining area (roughly 25,000 to 35,000 sq. ft.);

(3) three other ocean-front Black Point properties were previously subject to a two part valuation, 100% of benchmark for 10,000 sq. ft. and 50% for the balance — no evidence was presented as to the topography of these parcels;

(4) several Kahala Beach properties (thirteen) were subject to variable formulas in the past — some were assessed at 100% of benchmark for the first 10,000 sq. ft. with the balance at 50%, for others a three part valuation was used; however, the latter properties had been assessed at 100% of benchmark for 10,000 sq. ft. and 50% for an average of 26,000 sq. ft. (range — 10,000 to 40,000 sq. ft.) with relatively small remaining areas valued at 25% (average of 6,500 sq. ft., range — 1,800 to 14,000 sq. ft.);

(5) no specific evidence was presented as to the topography of the individual Kahala Beach properties, but the City assessor testified that (at least) some of them were level and had beach access — and, no distinction was made in valuing the overage of these lots as compared to those at Black Point.

II.

A. Grounds for Steiner's Tax Court Appeal

ROH § 8-12.3 provides that a taxpayer's assessment will be lowered if the taxpayer can show:

(1) assessment of the property exceeds by more than 10 percent the market value of the property, or (2) lack of uniformity or inequality, brought about by illegality of the methods used or error in the application of the methods to the property involved, or . . . (4) illegality, on any ground arising under the Constitution or laws of the United States or the laws of the state or the ordinances of the city in addition to the ground of illegality of the methods used, mentioned in clause (2).

ROH § 8-12.3 (1990). Steiner relied on all of these grounds in fashioning her appeal.

B. The Tax Appeal Court's Valuation

The Tax Appeal Court found that the City's 1988 and 1989 assessments: (1) exceeded the market value of the property by more than ten percent; (2) showed a lack of uniformity and inequality brought about by errors in the application of the methods used to the property involved; and (3) violated the equal protection clause of the U.S. Constitution. In addition to setting aside the City's valuation, the court calculated its own value for the property. The court applied the Kahala Beach benchmark value, rather than the Black Point benchmark, to 14,000 sq. ft. of the parcel and 25% of the Kahala Beach benchmark to the remaining 32,707 sq. ft., as follows: 1988

100% x $95 x 14,000 sq. ft. = $1,330,000 25% x $95 x 32,707 sq. ft. = 776,791 __________ $2,106,791 1989

100% x $125 x 14,000 sq. ft. = $1,750,000 25% x $125 x 32,707 sq. ft. = 1,022,094 __________ $2,722,094

III.

The Tax Appeal Court's valuation of the Steiner property requires us to review the three components which comprise the assessment. First, we must examine the benchmark value that was applied to the property. Second, we must consider the area of the property (in square feet) that was assessed at 100% of benchmark and the area that was assessed as overage. Third, we must decide whether the benchmark value was properly reduced and applied to the overage area of the Steiner parcel.

A. The Benchmark 1. Comparable Sales

The Tax Appeal Court, as part of its valuation of the Steiner property, rejected the City's benchmark value for ocean-front Black Point properties. The court substituted the Kahala Beach benchmark value. Although we presume that the court made this substitution in order to "arrive at a fair and realistic value" as we directed in Amfac, we find no evidence supporting the application of the Kahala Beach figure to the subject Black Point property. See In re Amfac, Inc., 65 Haw. 499, 502, 654 P.2d 363, 365 (1982). We have stated that "the tax appeal court . . . must base its conclusions upon evidence adduced and not upon what might have been adduced." In re Tax Appeals, 41 Haw. 141, 149 (1955) (quoting In re Taxes of Carter, 27 Haw. 826, 828 (1924)). Thus, we find that the substitution of the Kahala Beach benchmark for the Black Point benchmark was clearly erroneous.

In contrast, the City presented ample evidence to uphold the application of its Black Point benchmark to the Steiner property. The general method used by the City assessor in arriving at the benchmark comports with the market data approach required by ROH § 8-7.1(a) and described in the Procedure and Reference Manual. The assessor integrated the sales data for recently sold properties in the Black Point area into the benchmark. She also compared sales data from the Kahala Beach area, while recognizing that the Black Point benchmark should be lower because of the differences in the topographies of the two areas. Although the Procedure and Reference Manual is a department guideline, rather than a statute or ordinance, we find that compliance with the policies and procedures in the Manual is an indicia of fairness.

Tax valuation methods must be applied uniformly and equally, although not identically, to all taxpayers. See, e.g., Addington v. Board of County Comm'rs, 119 Kan. 528, 382 P.2d 315, 319 (1963) ("Uniformity in taxing implies equality in the basis of assessment as well as in the rate of taxation."); Kittery Elec. Light Co. v. Assessors, 219 A.2d 728, 734 (Me. 1966) (citations omitted) ("[U]niformity and equality in a constitutional and statutory sense does not require mathematical exactitude in the assessment valuation of property for taxation."). When the assessors follow department guidelines, there is a much greater likelihood that the tax burden will be equally shared by all taxpayers. See Allegheny Pittsburgh Coal Co. v. County Comm'n of Webster County, 488 U.S. 336, 345 (1989) (In finding an equal protection violation, the U.S. Supreme Court noted that the tax assessor's "practice seems to be contrary to that of the guide published by the West Virginia Tax Commission as an aid to local assessors in the assessment of real property."). Therefore, we hold that the City tax assessor's compliance with the policies and procedures set forth in the Finance Department's Procedure and Reference Manual is strong evidence that the City's method of mass valuation is being uniformly and fairly applied to all taxpayers.

2. Sales to Japanese Buyers

Steiner argued in the tax court that sales to Japanese investors were not "comparable" without adjustment, thus such sales should have been disregarded or discounted in calculating the benchmark value and the fair market value of Steiner's property. Steiner's experts opined that Japanese investors were paying more than fair market value because they had special influences, motivations and conditions and they were armed with "tremendous financial wallop." The experts testified that, at the time of the alleged "above market" sales, the dollar to yen ratio was high, Japan real estate prices were high, and low interest rates were available in Japan. Thus, they argued, Japanese buyers were paying much higher prices than were "economically justified" by any measure of value.

Although wealthy Japanese investors in Hawaii property may be a modern phenomenon, Steiner's theory is not. Bonbright's 1937 treatise on the appraisal of property discusses a similar theory that property should be valued at its "intrinsic value" or "justified price" rather than its fair market value. J. Bonbright, 1 VALUATION OF PROPERTY 24-29 (1937). Intrinsic value differs from fair market value "in that it represents, not what the property could presently be sold for, but what, in the appraiser's judgment, the property would sell for, were the market composed of (a) intelligent individuals who (b) were interested in buying and selling the property only by reference to its investment merits." Id. at 27. The intrinsic value theory mirrors Steiner's argument that sales to Japanese buyers are not economically justified, thus the court should adjust the sales data to reflect what would be a property's intrinsic value or justified price.

Bonbright states that one reason why courts have "wisely hesitated" to embrace the concept of intrinsic value is that "it would be extremely difficult to estimate." Id. at 28-29.

If [tax assessors or tax courts] were compelled to find the intrinsic value . . . for tax purposes, the administration of the law would become almost hopeless. Endless controversies would arise between taxpayers and assessors, many of which would be carried into the courts. Faulty as is market price as an index of value, it supplies a relatively objective and easily administered basis of valuation that no other method can supply.

Id. at 29. We agree with this analysis and will not inject uncertainty into Hawaii's property tax assessments by subscribing to Steiner's theory. We further note that Steiner's theory invites a sale-by-sale analysis of the "justified price;" we find this to be repugnant to the concept of "using appropriate systematic methods suitable for mass valuation." See ROH § 8-7.1(a). Thus, we reject Steiner's argument that sales to Japanese buyers do not reflect the fair market value of the properties sold to them.

In particular, we reject Steiner's argument that the 1987 sale of the parcel next to the Steiner property should be discounted because the buyer was a Japanese citizen and thus he paid more than fair market value. "Market value . . . is no more than the value in money of any property for which that property would sell on the open market by a willing seller to a willing buyer." In re Puna Sugar Co., 56 Haw. 621, 624, 547 P.2d 2, 4 (1976). The property next door was listed on the open market for 95 days at $3,000,000 and sold for $2,750,000; and, there is no evidence suggesting other than an arms-length transaction between a willing buyer and a willing seller. Furthermore, although the buyer was Japanese, he has owned and lived in a Kahala beach condominium since 1979 and is the owner of a local business, the Pearl City Tavern. Even if we agreed with Steiner's "above-market" theory, which we do not, nationality alone would be insufficient to establish that this transaction was tainted by the "Japan factor."

Although the Tax Appeal Court made "findings of fact" that recited Steiner's expert's testimony regarding the "above-market" theory, the court did not incorporate this theory into its own valuation of the subject property. The court merely substituted the Kahala Beach benchmark and applied it to a smaller portion of the Steiner parcel. The Kahala Beach benchmark was also based in large part on sales to "Japanese" buyers. Thus, the Kahala Beach benchmark used by the court was equally influenced by the alleged "above-market" sales and its use in no way indicates the court's adoption of Steiner's theory.

The City argues that the Tax Appeal Court disregarded sales to Japanese buyers in determining benchmark and fair market values, while also arguing that the court failed to adjust the Kahala Beach benchmark upward when the court applied them to less than 20,000 sq. ft. of the property. We find it equally plausible that the court used the Kahala Beach figure in an attempt to adjust upward from the Black Point benchmark when it decided to apply the benchmark to less than 20,000 sq. ft. Since both the Kahala Beach and the Black Point benchmarks incorporated sales to Japanese as well as other buyers, we find the City's argument on this point to be misguided. However, as stated above, we agree with the City that the tax court erred by using the Kahala Beach benchmark in this case. Rather, we conclude that the court should have applied the Black Point benchmark as the City did in its assessment.

B. The Overage Area

Steiner presented expert testimony from Jack Lipman, a licensed architect with over thirty years of experience, regarding the topography and usability of the land. Lipman testified that a little less than 11,000 sq. ft. of the Steiner parcel is buildable, except at extremely excessive costs. He mentioned "very rough and rugged and heavy rocks," inaccessibility for getting construction equipment in to do foundation work (pilings would be needed), the heavy winds and whip-lashing of water, and the Hawaiian trail across the property that cannot be built over.

The Tax Appeal Court found that 14,000 sq. ft. of the parcel were economically usable. The only (evidentiary) source of this information was the testimony of John Hulten, Sr., an appraiser called by Steiner. Mr. Hulten testified that he obtained the 14,000 sq. ft. figure, from an architect (Reese) who was not called as a witness in this case, to illustrate two appraisal approaches which he was advocating. Steiner's attorney stated that "as far as the report to Mr. Reese's report [sic] [on the buildable area], your honor, we're offering only to show what Mr. Hulten relied on but not to the truth of Mr. Reese's report."

As we have previously stated, an expert witness may not "serve as a mere conduit for the hearsay opinion, the factual basis of which is not established through evidence, of another expert who does not testify when the expert who does testify lacks the requisite qualifications to render the opinion in his own right." State v. Davis, 53 Haw. 582, 589-90, 499 P.2d 663, 669 (1972) (citations omitted). In this case, no attempt was made to establish Hulten's qualifications to determine the usable area of the Steiner property and the factual basis for the 14,000 sq. ft. figure was not otherwise established through evidence presented by either party. We therefore conclude that the Tax Appeal Court erred in finding 14,000 sq. ft. of the Steiner property usable for improvements.

The City attacks the 14,000 sq. ft. figure from another perspective. It argues, correctly, that a benchmark value is calculated for a "typical" or benchmark lot size and, when a property is smaller than the typical lot, the benchmark value must be adjusted upward to reflect greater value per square foot. Steiner does not disagree with this analysis. In fact, Steiner's expert diagramed a hyperbolic curve of the formula for calculating the adjustment ratios to be applied to lots smaller than 20,000 sq. ft.

We agree that this analysis would be applicable to lots smaller than the benchmark lot. However, Steiner's property is actually much larger than the benchmark or typical lot. We will not create an anomaly whereby a single parcel is viewed as being both larger and smaller than the benchmark lot.

Steiner presented evidence that the usable area of the property was significantly less than 20,000 sq. ft. Nevertheless, the City's study showed that a typical lot size along the ocean-front in both the Black Point and Kahala Beach neighborhoods was 20,000 sq. ft. The City assessor testified that the non-utility of a portion of the benchmark area was factored into the calculation of the benchmark value. We find that application of the Black Point benchmark value to 20,000 sq. ft. is an "appropriate systematic method" for valuation of this portion of the Steiner property. See ROH § 8-7.1(a). We hold that the tax court erroneously rejected the City's valuation of the parcel using a formula which applied 100% of the Black Point benchmark value to 20,000 sq. ft. and a reduced benchmark to the overage area of 26,707 sq. ft.

C. The Reduced Benchmark Value as Applied to Overage

The Tax Appeal Court valuation applied 25% of the benchmark value to the Steiner overage area. We cannot conclude that there is "good and sufficient reason" to overturn this decision. See In re Puna Sugar Co., 56 Haw. 621, 623, 547 P.2d 2, 4 (1976). The court's conclusions of law indicate that the City failed to adequately adjust for dissimilarities between Kahala Beach overage and Black Point overage, including topography, shoreline access, utility and other factors affecting valuation.

While we found that the City's benchmark sufficiently incorporated these factors, we agree with the tax court that the overage calculation did not. Contrary to its eighteen year practice which was based on the characteristics of the overage area of the Steiner property, the City used the same overage formula for unbuildable, rock-strewn overage with no beach access as it did for level, usable Kahala Beach beach-front overage. The City failed to adjust its overage formula in contradiction of its own assessment guidelines. See Procedure and Reference Manual §§ 911.00, 913.00, 941.31 941.34; see also Part III. A. 1., above. The court concluded that the City's assessment discriminated against the landowner by requiring payment of more than the land's allocable share of the total real property tax burden, thus violating equal protection. We agree.

The equal protection clause of the U.S. Constitution is violated when taxation "in fact bears unequally on persons or property of the same class." Allegheny Pittsburgh Coal Co. v. County Comm'n of Webster County, 488 U.S. 336, 343 (1989) (citations omitted); see also In re Swann, 7 Haw. App. 390, 401, 776 P.2d 395, 401-02 (1989). "[T]he fairness of one's allocable share of the total property tax burden can only be meaningfully evaluated by comparison with the share of others similarly situated relative to their property holdings." 488 U.S. at 346. In this case, fairness requires the assessment of Steiner's overage at a lower percentage of benchmark than Kahala Beach overage, because each has significantly different characteristics affecting value. By failing to reduce the overage assessment, the City denied Steiner's right to equal protection under the law.

IV.

In summary, we reverse in part, holding that the Tax Appeal Court erroneously set aside the City's application of the Black Point benchmark value to 20,000 sq. ft. of the Steiner property. We affirm in part, holding that the court correctly applied the 25% of benchmark to overage in lieu of the 50% of benchmark used in the City's formula. The assessed value of the Steiner property must be computed as follows:

1988

100% x $75 x 20,000 sq. ft. = $1,500,000 25% x $75 x 26,707 sq. ft. = 500,756 __________ $2,000,756

100% x $95 x 20,000 sq. ft. = $1,900,000 25% x $95 x 26,707 sq. ft. = 634,292 __________ $2,534,293

We affirm the tax court's ruling that the 1988 and 1989 assessments for the Steiner parcel: (1) exceeded the market value of the property by more than 10%; (2) showed a lack of uniformity and inequality brought about by errors in the application of the methods used to the property involved; and (3) failed to provide Steiner equal protection under the law.

Accordingly, the judgment of the tax court is affirmed in part and reversed in part. We remand for entry of judgment consistent with this opinion.