[42]*42OPINION OP THE COURT
BY WIRTZ, J.
This case involves real property taxes for Ewa Plantation Company and Waialua Agricultural Company, Limited, for tbe year 1956. The total tax for each taxpayer was computed on the basis of the value of an average acre of cane land. The taxpayers did not object to the method whereby the value of an average acre of cane land was used to determine the value of all of their cane land for assessment purposes. Rather, the dispute centered on the method used by the Tax Commissioner of the Territory of Hawaii1 to determine the value of an average acre of cane land and to the values obtained by that method. The sole question on appeal, as it was in the tax appeal court, is the value of an average acre of cane land of each of the taxpayers.
[43]*43In 1955 the tax commissioner formed a committee to study and recommend fair and reasonable values of sugar cane lands for the assessment of real property taxes in 1956. It had been the practice to assess sugar cane lands every four years and the previous assessment had been made in 1952. The formula and the values submitted by the committee were adopted by the tax commissioner in making the assessments from which the taxpayers appealed to the tax appeal court. The values forming the basis of this assessment for 1956 were $735 for Ewa Plantation Company and $658 for Waialua Agricultural Company, Limited.
The taxpayers’ position was that the 1952 values plus an increase of ten per cent. (10%) represented a fair and reasonable value. Under the taxpayers’ method of valuation, the resultant values were $635 for Ewa Plantation Company and $576 for Waialua Agricultural Company, Limited.
The tax appeal court, in rejecting the values derived from the formula of the committee adopted by the tax commissioner, concluded that the values produced by the tax commissioner’s formula were not accurate and should not be used for taxpayers’ lands. It then proceeded to value taxpayers’ sugar cane lands by engrafting on the method advocated by the taxpayers an additional five per cent. (5 %) on the 1952 values to compensate for the island differential in values. This resulted in a value of $664.30 for Ewa Plantation Company and $603.75 for Waialua Agricultural Company, Limited. Prom this decision of the tax appeal court, entered on April 10, 1959, the tax commissioner has brought this appeal.
It is well established that upon an appeal from the tax appeal court, the appellant has the burden of establishing that the decision of that court was erroneous. Tax Appeals, Maenaka, 41 Haw. 141; In re Taxes Castle, 24 [44]*44Haw. 598; In re Taxes Waiakea Mill Co., 24 Haw. 333; In re Taxes, Catholic Mission, 22 Haw. 764; Hawi Mill & Plantation Co. v. Forrest, 21 Haw. 389; In re Taxes, Makee Sugar Co., 19 Haw. 331; Tax Assessor v. Wailuku Sugar Co., 18 Haw. 422; Tax Assessor v. Wilder, 17 Haw. 425; In re Taxes, Haw’n Sugar Co., 16 Haw. 236; Lihue Plantation Co. v. Farley, 13 Haw. 283. This court, in Tax Assessor v. Wailuku Sugar Co., supra, at page 423, stated the rule thusly:
“This court has uniformly held that it does not reduce or increase the valuation made by a tax appeal court which appears to be fair and just, but allows it to stand unless shown to be erroneous, or based on a wrong theory or insufficient or defective data. Hind v. Willfong, 13 Haw. 125; Assessor v. Rapid Transit Co., 15 Ib. 3; Oahu R. & L. Co. v. Assessor, 17 Ib. 163; Tax Assessor v. Wilder, Ib. 425.”
Again, the court in In re Taxes Bishop Est., 33 Haw. 149, 159, compared the findings of the tax appeal court with those of a circuit judge by stating:
“* * * The court has on numerous occasions announced the rule to be that the findings of a tax appeal court are entitled to great weight; that where such findings depend upon the credibility of witnesses and upon the weight of conflicting statements of witnesses, such findings are to be accorded the same weight as the findings of a circuit judge at chambers. (See Tax Assessment Appeals, 11 Haw. 235; Hawi Mill & Plantation Co. v. Forrest, 21 Haw. 389.) In the Hawi Mill & Plantation Co. case this court expressly likened tax appeals to equity appeals.
“* * * Indeed, under the new law the tax appeal court has been raised to the dignity and importance of a court of record, clothed within the sphere of its duties and functions with all the power and authority [45]*45of a circuit judge at chambers. (See §§ 43, 48, Act 40, 2d Sp. S. L. 1932.2)”
Consequently, we approach the first question presented under this appeal, namely, whether the tax appeal court erred in its determination of the respective values of an average acre of cane land for each of the taxpayers, with the admonition that such can only be disturbed if found to be “clearly erroneous.” H.R.C.P., Rule 52(a); Filipino Federation of America, Inc. v. Cubico, 46 Haw. 353, 380 P.2d 488; Peine v. Murphy, 46 Haw. 233, 377 P.2d 708; Miller v. Loo, 43 Haw. 76; Lum v. Stevens, 42 Haw. 286; Hawaii Builders Supply Co. v. Kaneta, 42 Haw. 111; Lima v. Tomasa, 42 Haw. 478. In this connection we must keep in mind the warning of In re Taxes Carter, 27 Haw. 826, 828, that “* * * the tax appeal court, similarly as this court, must base its conclusions upon the evidence adduced and not upon what might have been adduced
The tax appeal court’s method of redetermining the values of an average acre of each of taxpayers’ cane lands may be briefly summarized. First, it noted that the 1952 values, having been accepted by both the tax commissioner and the taxpayers, were the only figures having any real value. This became the foundation for its revaluation. The court, having noted that the values of agricultural lands appeared generally to have increased about ten per cent. (10%) from 1952 to 1956, concluded that the ten per cent. (10%) increase testified to by taxpayers’ expert witness was reasonable. Finally, the court went on to find that the values of lands on the island of Oahu were subject to special influences which warranted an additional five per cent. (5%) increase over 1952 values. In short, the court revalued an average acre of taxpayers’ [46]*46cane land at fifteen per cent. (15%) over the 1952 values.
The tax appeal court’s reliance, in this instance, on the 1952 values as the sole basis from which to compute, percentagewise, the values for 1956 is not in itself a sound method of approaching an assessment.3 In Hackensack Water Go. v. Division of Tase Appeals, 2 N.J. 157, 65 A.2d 828, 830, it was pointed out that:
“* * * Each annual assessment of property for taxation is a separate entity, distinct from the assessment of the previous or subsequent year * * *.
“On each assessment the test is does it reflect true value?”
Again, the court in In re Kresge-Newark, Inc., 30 N.J. Super. 489, 105 A.2d 12, 18, emphasized that:
“* * * Such assessment for prior years should not be used as the sole basis of valuation for the year in question because such assessment for prior years may not reflect true value * * (Emphasis added.)
If reliance on prior assessed values is unsound when such values are based on bona fide opinions of values derived from appraisal, it is even more unsound where [47]*47such values are obtained through negotiation as was the case here. The record discloses a compilation of the average estimate of value per acre covering sugar cane lands of plantations on the island of Oahu for the year 1952. Therein enumerated were values found by eleven qualified real estate appraisers ranging from $525 to $750 per acre in the cases of Ewa and Waialua. The average per acre value attributable to Ewa was $611 while that attributable to Waialua was $585. These average values were then adjusted to $578 for Ewa and $525 for Waialua, as the basis for the 1952 assessment, by negotiation. Since the foregoing questionable procedure of bargaining had been applied to determine the agreed valuation of 1952, it follows that there exists even a more serious question as to the reliability of these 1952 values which the tax appeal court used as its sole foundation for the 1956 values.
The taxpayers assert that: “This is exactly what was done in Tax Appeals, Maenaka, [supra] 41 Haw. 141 (1955), where this court sustained the tax appeal court’s determination that values for 1949 be computed on the basis of 1948 values plus 20%.” While this court in Maenaka, supra, did sustain the conclusion of the tax appeal court, it did not do so on the sole basis of a percentagewise increase of the 1948 assessments as it stated:
“In view of the mandate of Section 5146, Revised Laws of Hawaii 1945, the record is replete with evidence bearing upon the advantage or disadvantage of location, accessibility, transportation facilities, size, shape, topography, quality of soil, water privileges, availability of water and its cost, productivity and nature of use, and to the opinions of persons who may be considered to have special knowledge of land values, and further to all other influences whether similar to those listed in the statute or not, which fairly and [48]*48reasonably bear upon the question of value.
“We are cognizant that this court is not limited by the findings of fact as determined below, and may probe the record in order to weigh the evidence or permit introduction of additional evidence material to the matter in dispute. (R.L.H. 1945, § 5214; In Re Leo L. Yerian, 35 Haw. 855.) Reference to or discussion herein of the voluminous evidence presented by the parties which was considered in the determination of land value would serve no useful purpose save to illustrate that the tax appeal court did in fact have ample competent evidence before it and correctly concluded as it did. We find no reason to disturb that conclusion.” (41 Haw. 154.)
Further, while there was evidence in the record to support a finding that in 1956 cane land values had increased approximately ten per cent. {10%) over those of 1952, still we can find no evidence in the record to substantiate the additional increase of five per cent. (5%) made by the tax appeal court in recognition of an island differential for cane land values. There was only the evidence that, in the opinion of the sugar committee as adopted by the tax commissioner, a nine per cent. (9%) over-all average effect was considered reasonable for use in connection with a formula based on certain standardized earnings and costs. If the 1952 values had been actually based on real estate appraisals there would be no necessity for an island differential as such would necessarily be reflected in those values. It is only when a formula, based on average and standardized factors applicable to all plantations throughout the state, is utilized that a need might arise to make an adjustment in the resultant land values in view of the greater competitive use of land on the island of Oahu.
Having seen that the tax appeal court’s method of [49]*49revaluing the average acre of taxpayers’ cane lands was founded on the shaky foundation of assumed values adjusted arbitrarily without evidentiary support to compensate for a supposed island differential, we can only conclude that the tax appeal court’s conclusions of the 1956 values for taxpayers’ sugar cane lands were “clearly erroneous” and must be set aside.
We, then, deem it necessary to focus our attention on the values, upon which the tax commissioner premised his 1956 assessment, found to be invalid by the tax appeal court. This brings us to the remaining question under this appeal. Did the tax appeal court err in determining that the values produced by the tax commissioner’s formula were not accurate and should not have been used for taxpayers’ lands?
In support of its conclusion that the formula did not give true values and should not have been used for the taxpayers in this case, the tax appeal court made the following findings: (1) that the discrepancy between the values of an acre of land of Oahu Sugar Company and EAva Plantation Company was “entirely too great”; (2) that the formula assigns value to plantations as going concerns rather than to the plantations’ cane lands themselves; (3) that the capitalization of rentals was inaccurate because, among other things, whereas the rentals were shown to be approximately half basic rental — normally capitalized at five per cent. (5%) — and half percentage rental — normally capitalized at ten per cent. (10%) — the total rentals were capitalized at five and one-third per cent. (5-1/3%); and (4) that the capitalization of computed rentals was geared to net earnings and produced figures that were out of line.
It should be pointed out at the outset that even if the Oahu Sugar Company assessment is on the low side of [50]*50actual market value, as the testimony shows,4 this fact, of and by itself, provides no relief for the taxpayers. An erroneous assessment is not established merely because other property has been assessed too low if the tax commissioner has acted in good faith and if other property in general has not been assessed at a lower rate. In re Taxes Onomea Sugar Co., 25 Haw. 278, 283; Re Taxes Menefoglio, 25 Haw. 106-107; and Re Taxes Swain, 29 Haw. 183-184, in which this court said:
“It has been repeatedly held by this court that the issue in any given tax case is whether the property involved in the appeal was assessed too high and not whether other property was assessed too low. ‘The question is whether the appellant’s property is assessed too high, not whether some other properties are assessed too low. If it appeared that other properties generally were assessed at a lower rate, it might be proper to assess this at the same rate * * *.’ ”
However, such grossly low assessment considered in conjunction with other erroneous or discriminating factors might well tend to establish the higher disproportionate valuation as unfair and unreasonable.
In attacking the 1956 assessment, we are mindful that the taxpayers have the burden to show clearly the invalidity claimed by overcoming the presumption that the tax assessor has faithfully performed his duty. Weyerhaeuser Land Co. v. Board of Equalization, 85 Or. 434, 165 Pac. 1164; Knappton Towboat Co. v. Chambers, 202 Or. 618, 276 P.2d 425, as modified by 207 Or. 702, 277 P.2d 763; Bailey v. Megan (8th Cir.), 102 F.2d 651. Cf., R.L.H. 1945, § 5127, as amended (now R.L.H. 1955, § 115-25). In People ex rel. Rickey v. Hunt, 241 App. [51]*51Div. 261, 271 N.Y. Supp. 842, 846, it was stated: “* * * that a party who assails the validity of an assessment must make it conclusively appear that the method by which the assessors arrived at the result complained of was incorrect, and that the assessment does not represent the fair value of the property assessed * *
However, this presumption, being an evidentiary presumption, merges with the evidence presented relating to the presumed fact, serving its function by reminding the trial court (here the tax appeal court) of its task to weigh this evidence with due regard to the burden of proof thereby imposed on the taxpayers. Such function is limited to the tax appeal court. In reviewing the decision and findings of that court, as we have seen, the presumption now arises favoring its actions; that its findings and decision are entitled to substantial weight and should not be overturned without good and sufficient reason and that the appellant has the burden of showing that the decision of the tax appeal court was “clearly erroneous” in finding that the taxpayers sustained their burden of overcoming the presumption favoring the 1956 assessment.
With this in mind, we turn to a consideration of the legislative requirements governing the duties and functions of the tax commissioner and which outline and control his performance in the valuation of the real property within the state. The pertinent sections of the Revised Laws of Hawaii 1945, as amended, provide that the objectives of the tax commissioner (now director of taxation) are to determine “fair and reasonable value” (R.L.H. 1945, § 5139, now R.L.H. 1955, § 128-2) and to do so “by appropriate systematic methods so selected and applied as to secure, as far as possible, uniform and equalized results throughout the [state]” (R.L.H. 1945, § 5146, as amended, now R.L.H. 1955, § 128-9). He is authorized to obtain “the opinions of persons who may be considered to have [52]*52special knowledge of land values” (R.L.H. 1945, § 5146, as amended, now R.L.H. 1955, § 128-9) and is required, so far as practicable, to keep records “which shall show the methods established” (R.L.H. 1945, § 5146, as amended, now R.L.H. 1955, § 128-9).
The tax commissioner’s committee of 1955 was organized, with persons possessing special knowledge of land values included in its membership, to study and recommend methods to obtain fair and reasonable values of sugar cane lands for purposes of the real property tax for the 1956 assessment. Representatives of the five sugar agencies, three independent real estate appraisers, the tax assessor (who served as chairman) and the tax research analyst, comprised the committee. The makeup of the committee as a whole, or the qualifications of any of the individual members, has not been challenged by the taxpayers. After extensive study of the problem of valuing cane lands in the absence of comparable sales data, the committee adopted a complex and involved formula which, from data supplied by the particular sugar plantations and inserted therein, produced values supposedly representing the values of an average acre of cane land for each plantation. The report of the committee, adopting such involved formula, was approved by all the members except Mr. Kenneth Nurse, the representative of Castle & Cooke, Ltd., agents for the taxpayers among others. This report here in evidence, was accepted and adopted by the tax commissioner.
Turning now to the formula, it is made up of five methods of valuation which were assigned weights to be used in combining the five components for the determination of a mean value upon which the 1956 assessment was based. The first method was the value of an average acre of cane land determined by the so-called “V/U” method. It was designated the land unit value and as[53]*53signed a weight of 4 points out of 20 (20 per cent.). The second was the value based on the capitalization of actual rentals and carried a weight of 1 point (5 per cent.). The third was the value based on the capitalization of computed rentals and commanded a weight of 2 points (10 per cent.). The fourth was the value used in the 1952 assessment which bore the greatest weight of 10 points (50 per cent.). The final method was the value determined according to productivity and was given a weight of 3 points (15 per cent.).
We now thread our way through the maze of the formula by considering each method therein individually. In so doing, we do not profess to have a complete understanding of the intricacies of the technical workings of the formula. We note, however, that we do not stand alone in our inability to grasp all the niceties and mathematical nuances contained in the formula, for it is evident from the record that such inability was shared by all of the expert witnesses appearing in the tax appeal court, some of whom were members of the committee and including the deputy tax commissioner charged with the responsibility for real property assessments on the island of Oahu, with the possible exception of the alleged father of the most involved portion of the formula, the V/U method, Mr. Joel Cox, who nevertheless in turn had difficulty explaining it to unresponsive minds.
We proceed then to consider first the V/U or land unit valuation method utilized under the formula. This method of determining value involves several steps based on complicated procedures of statistical analysis. Mr. Cox, in his testimony, described this method as having the purpose of freeing an income approach to market land value from as many as possible of the accidental short term variables caused by temporary conditions, management practices and extraneous influences. Eeduced to its sim[54]*54plest terms it appears that the V/U method is intended to work as follows: A determination is made of the standard value of one ton of sugar at the mill for each individual plantation. From this value are subtracted the actual costs incurred by the particular plantation in producing that ton of sugar. The difference is the profit, a portion of which is attributed to rent which is then capitalized to give the value of the land on which the sugar cane was grown.
In determining the value of one ton of sugar at the mill, the starting point is the average profit per ton for the industry as a whole during previous years. To this profit is added the cost of production as determined by the records of the industry as a whole. It is assumed that this cost will vary according to the size of the plantation, that is, the cost of producing one ton of sugar will be less on a larger plantation than on a smaller one. Consideration is also given to the varying distances the sugar must be transported for shipment. The resultant differences in the value of one ton of sugar to each plantation are, however, minimal.
The second step in the V/U method is to ascertain the profit made by each plantation on the aforementioned ton of sugar. This is determined by subtracting from the value of the ton of sugar at the mill the actual costs of production of the particular plantation under consideration. These costs are not industry-wide averages, but, instead, are the actual costs incurred by the individual plantations. The purpose of using actual field costs in this income approach, according to Mr. Cox, was that no other suitable method at that time had been developed. These differences in actual costs were substantial, resulting in marked differences in the values of the lands of the various plantations as determined by the V/U method. Mr. Cox admitted that the cost figures for each plantation [55]*55were influenced by management as well as by the physical characteristics of the land and that it was impossible to separate the effects of these two factors. The tax commissioner concedes “that no method has as yet been developed to eliminate entirely the influence of management in the income approach so [that] this factor may be distinguished from the effect of physical characteristics on each plantation.”
The third step under the V/U method was to determine the profit per standard ton of sugar by subtracting cost from value. Finally, this profit per ton of sugar was converted into a figure supposedly representing the value of the land. This was accomplished by attributing a portion of the profit to rental for the land. This rental was then divided into a base rental capitalized at five per cent. (5%) and an additional rental capitalized at ten percent. (10%). The sum of these two figures represented the value according to the V/U method.
These values were modified by a so-called “island differential.” This additional step was obtained by capitalizing the “residual earnings” per acre of each plantation.5 The tax committee then found that on the average the values determined in this fashion were 1.301 times larger than the values determined by the V/U method, giving “corrected” V/U values, which were used in the formula in setting the final values for the 1956 assessments.
It is necessary to consider this so-called “island dif[56]*56ferential”6 in more detail. It was supposedly used to provide relativity between values on the island of Oabu and those on the neighbor islands. It caused an over-all average assessment increase of approximately nine per cent. (9%) for the sugar cane lands on the island of Oahu and this percentage difference was designed to reflect the effect of competition on land values on the island of Oahu.
In the first place, Mr. Cox admitted that there was no logical basis for the use of the “residual earnings” method to create an “island differential” factor.7
In the second place, Mr. Cox admitted in his testimony that no attempt had been made to evaluate the effect of competing uses on the island of Oahu. As to what uses he had in mind, he testified that he was thinking primarily of pineapples. However, he admitted that this would not affect the lands of Ewa Plantation Company.8
Finally, a consideration of the mathematical steps [57]*57taken in applying the so-called “island differential” to arrive at the final or corrected Y/U values shows that the original Y/U values were not merely modified, hut were actually eliminated from the method entirely and in their place were substituted the so-called corrected Y/U values, which were actually based only on residual earnings. This is so because the ratio used, which was determined by dividing the residual earnings values by the original Y/U values, in the so-called correction would vary proportionately to the original Y/U values so that the so-called corrected Y/U values would remain constant regardless of the values originally assigned under the Y/U method. That the use of the residual earnings method is not an appropriate appraisal method is clearly acknowledged by Mr. Cox in his testimony.9
We consider it appropriate in closing our analysis of the Y/U or land unit method of valuation utilized under the formula to quote from a survey report made to the tax commissioner by the Public Administration Service on December 1, 1958, entitled Real Property Assessment in Hawaii, supra, which is an exhibit in evidence in the record, as follows:
“The land unit valuation method * * * may be one of the most highly ingenious approaches ever made to the problem of determining value. Through the use [58]*58of a battery of constants and variables incorporated in a series of complex mathematical relationships this method attempts to determine land value and island differentials independent of the level of plantation management. It is almost immaterial whether the formula succeeds in its objective or not, because nothing can justify use of a method that the assessor does not understand, that most taxpayers cannot understand, and that cannot be explained satisfactorily to anyone legitimately interested.” (pp. 118:119.)
In general, the method based upon the capitalization of actual rentals utilized under the tax commissioner’s formula is vulnerable to attack from the failure to scrutinize the terms of each lease and the conditions under which each lease was executed as admitted by Mr. Cox in his testimony.10 Yet, Mr. Cox testified that in “any appraisal based on the land on which the rentals are obtained is to be made, a very careful study should be made, an analysis of all circumstances surrounding the lease.” It was admitted that failure to do so could and did produce “erratic [59]*59results in respect to” individual plantations.11 Further, the rental figures used pertained only to rentals paid under leases for areas in excess of 500 acres and gave an incomplete and inaccurate picture of the over-all average of the rent paid by particular plantations.
The manner and rate of capitalization employed under this method, in itself, presents a distorted picture of values. According to the testimony the accepted way of capitalizing rental which included both a minimum and a differential (based on percentage of income) rental was to capitalize the minimum rental at five per cent. (5%) and the differential rental at ten per cent. (10%). In spite of this, however, the manner and rate employed provided for the capitalization of total rents, both minimum and differential, at five and one-third per cent. (5-1/3%). In the case of the taxpayers, whose leases included minimum rentals and differential rentals in about equal proportions, this resulted in a variance in values of from twenty-six per cent. (26%) to thirty-four per cent. (31%).
Also, under this method leases were included regardless of when they Avere negotiated. It is fundamental that “an old lease is just as unsatisfactory an indicator of current value as an old sale.”12
The method of determining values based on the capitalization of computed rentals was evolved and used as a separate approach, according to Mr. Cox, because the actual rental method valuations were erratic. This was a statistical method based upon actual rentals and net [60]*60earnings for the various plantations. It approximated the actual rentals by a comparison with earnings per acre in which the rental was expressed as a percentage of net earnings or profit. “The relationship of earnings to the percentage of earnings actually paid out in rent was used in calculating computed rentals, or what formerly was called the landlord’s share for irrigated plantations.”13 The theory behind this method was that the percentage decreased as the profit increased. Using this data, a mean curve or relationship between the two figures was developed and a value per acre determined. However, the wide variation (as high as twelve per cent. (12%)) between the actual percentage and the computed percentage of earnings attributable to rent tends to discredit the results obtained. Again, the manner of capitalization of the computed rentals involved the use of a constant rate of five and one-third per cent. (5-1/3%) instead of the customary and accepted practice of capitalizing the minimum rental at five per cent. (5%) and the differential rental at ten per cent. (10%), subjects this procedure to the criticism already made of its use in the method of the capitalization of actual rentals.
The fourth method employed by the tax commissioner under the formula was the use of the 1952 assessed values. What has already been said about the reliance on past values to compute present values as being unsound applies equally here. Under the formula, the 1952 assessed values were given the greatest weight, 10 points or fifty per cent. (50%). Actually it carried a greater weight as these 1952 assessed values were also incorporated in the final method next to be considered.
The final method in the tax commissioner’s formula was that by which values were determined by productivity [61]*61of the various plantations. This method is based on the actual data pertaining to productivity for each plantation for the five preceding years. The figure used for each plantation was the average number of tons of sugar produced per acre per month.
It is true, as testified to by Mr. Cox, that these figures depend on the character of the land involved. But only to the extent that farming practices are uniform. Land that is better situated in respect to sunlight and salt-free water will grow more cane than land that is less favorably situated in these respects. However, it is equally true that the amount of sugar cane grown will also be affected by the amount of water and fertilizer that management puts on the land and by the variety of cane that management has planted. Since the productivity figure relates to the average amount of sugar produced per acre and not to the amount of sugar cane grown, it is affected, as alluded to under the criticism of the Y/U method in connection with actual production costs, not only by the physical characteristics of the land and the effects of management as they relate to growing cane, but also by the effects of management in preparing the cane for harvest, harvesting and grinding the cane and extracting the sugar therefrom. It is seen, then, that the heavy hand of management is felt throughout the formula in its effect on the resultant values obtained.
A consideration of the details of the productivity method discloses that it, likewise, is a complicated one. The productivity for each plantation in terms of tons of sugar per acre per month was plotted on a chart against the value of an average acre of cane land for each plantation as determined by the other methods of valuation utilized in the formula. A curve representing an average of these plotted points was then drawn through these plotted points giving the values, which in turn were ad[62]*62justed by the use of the “island differential” factor.
It is clear, then, that all of the errors inherent in the other methods of valuation, above considered, were repeated when the values produced by these other methods were incorporated into the productivity method of valuation, thus increasing the effects of these errors on the final values determined under the formula.14
Again, we find confirmation of our criticism in Real Property Assessment in Hawaii, supra, where it is stated at page 119:
“The * * * approach is based on productivity by plotting production against value per acre for each irrigated plantation. The value used, however, is computed by averaging the 1952 assessment with the values determined using some of the 1955 measures. This hardly represents an independent measure of value but is primarily a way of expressing values already determined in a little different way. The heavy weight given the 1952 assessment in figuring this particular series of values is in addition to the 50 per cent weight assigned to it in figuring the final average per cent assessed value.”
The foregoing brief and limited analysis of the formula used by the tax commissioner for the 1956 assessment, shallow and inadequate- as it may seem due to an incomplete understanding of its intricacies, reveals many glaring defects which cast a grave doubt on the accuracy of the results obtained therefrom. The tax commissioner contends “the formula is complex and it is the judgment areas therein which the taxpayers have challenged. The [63]*63tax commissioner’s position is that these criticisms do not justify a finding that the method used was illegal,” citing Bailey v. Megan, supra (8th Cir.), 102 F.2d 651; Weyerhaeuser Land Co. v. Board of Equalization, supra, 85 Or. 434, 165 Pac. 1164; People ex rel. Rickey v. Hunt, supra, 241 App. Div. 261, 271 N.Y. Supp. 842 and Knappton. Towboat Co. v. Chambers, supra, 202 Or. 618, 276 P.2d 425, as modified by 207 Or. 702, 277 P.2d 763. From the record it is clear that the committee combined several inflexible appraisal methods to create a single inflexible formula. Data such as field costs, lease rentals and the like for the various plantations separately was inserted in the formula and the resulting answers were taken as the market values used for the respective plantations. Any judgment exercised by the committee was in respect to the adoption of an inflexible formula. Once the formula had been adopted, judgment played no part whatsoever in applying the formula to the respective taxpayers or in respect to the final results produced by the formula. The tax commissioner in turn exercised no judgment whatsoever with respect to the final formula but adopted it and its results; without modification. The application of the formula with assigned weights to the various methods involved obviated the use of judgment on the part of the tax commissioner and necessarily obscured the subjective nature of an appraisal.
Were we satisfied that the results obtained from this formula reflected land values as such, as well as land values which pertain to the current assessment year, we might be faced with the rule that in addition to showing conclusively that the method employed was incorrect, the taxpayers are nonetheless required to show that the results incorporated in the assessment do not represent the fair and reasonable value of the land assessed. Bailey v. Megan, supra (8th Cir.), 102 F.2d 651; Weyerhaeuser [64]*64Land Co. v. Board of Equalization, supra, 85 Or. 434, 165 Pac. 1164; People ex rel. Rickey v. Hunt, supra, 241 App. Div. 261, 271 N.Y. Supp. 842; Knappton Towboat Co. v. Chambers, supra, 202 Or. 618, 276 P.2d 425, as modified by 207 Or. 702, 277 P.2d 763; In re Kresge-Newark, Inc., supra, 30 N.J. Super. 489, 105 A.2d 12. Cf., Oahu R. & L. Co. v. Assessor, 17 Haw. 163.15 However, from our analysis of the formula we entertain grave doubts that the results obtained truly reflect land values and that the formula employed is an “appropriate” method required by R.L.H. 1955, § 128-9 although it was admittedly “systematic.” As the formula is based on present use without regard to other property subject to assessment, it seems to reflect values attributable to the use of land instead of the value of the land itself.
Consequently we are loath to disturb, as “clearly erroneous,” the tax appeal court’s conclusion that the taxpayers had sustained their burden in overcoming the presumption favoring the 1956 assessment, and among other things, “that the tax assessor’s formula, regardless of what attempts had been made, actually was a formula which was valuing a plantation as a going concern rather than the cane lands and acreage in cane as such alone,” and again that it “was not accurate as to the actual valuation of cane land as cane land but, actually was valuing the plantations’ operation.”
In Real Property Assessment in Hawaii, supra, at pages 119-120, appears the following obituary of the formula: “The most sympathetic student of the present method of valuing cane land would have to admit that it is theoretically indefensible from the viewpoint of the as[65]*65sessor. If it produces equitable results, as it appears to do, they must be attributed to the committee members’ deep knowledge of cane land values, to their ingenuity in devising additional factors and modifiers when the results of one approach were incorrect, to the imagination with which they viewed their scatter diagrams, to the accuracy of values which had been derived by somewhat similar methods in 1952, and to the sense of responsibility and restraint which committee members have exercised in developing means of assessing their own industry.”
While this would seem to indicate that the results were equitable it should be noted that from the record this would appear to be so in case of all of the plantations involved with the exception of the taxpayers. The difference assigned to the values of the taxpayers’ lands in comparison with the other plantations situated on the island of Oahu under the 1956 assessment were way out of line and disproportionate to those assigned in the 1952 assessment. Especially was this true in connection with the lands of Oahu Sugar Company, some of whose cane fields were adjacent to those of Ewa Plantation Company. The only explanation given for this discrepancy as to Oahu Sugar Company was the loss of a large portion of its most productive land which resulted in that plantation being less productive as a unit and thus reflecting a sharp downward trend in the valuation of its sugar lands. This explanation alone tends to support the tax appeal court’s conclusion that the formula valued “a going concern rather than the cane lands” and emphasized the effects of management on the results obtained under the formula.
While the fact that the lands of another are undervalued alone does not justify the setting aside of an assessment, as seen above, this factor coupled with the grave inaccuracies revealed in the methods employed by the tax [66]*66commissioner tends to emphasize the inappropriateness of the formula for valuation purposes. The tax appeal court found that “this discrepancy [between the lands of Oahu Sugar Company and Ewa Plantation Company] was entirely too great” and concluded that “the results obtained from the use of the formula were obviously inaccurate.”
R. B. Greig, Deputy Attorney General (also on the briefs), for appellant Tax Commissioner.
R. E. Btifel (Anderson, Wrenn, & Jenks with him on the brief) for appellees.
We can reach no other conclusion than that the formula used by the tax commissioner in arriving at the 1956 assessments was not appropriate. Rather it was inappropriate and illegal and involved an erroneous application of appraisal methods to the taxpayers’ lands.
Having determined that the values fixed by the tax appeal court to have been “clearly erroneous” and that the 1956 assessments based on the tax commissioner’s formula were correctly rejected, thus removing from consideration any evidence that might form the basis for a legal assessment of taxpayers’ lands for 1956 by the tax appeal court, we have no alternative, but to remand the matter to the tax appeal court with instructions to, in turn, remand the same to the tax commissioner for reassessment. Reversed in part and affirmed in part.