The opinion of the court was delivered by
Francis, J. A. D.
The Boroughs of Carteret and Sayreville and the Township of Woodbridge appeal from the August 30, 1955 judgment of the Division of Tax Appeals entered following its hearing for the purpose of revising and correcting the equalization table adopted by the Middle-sex County Tax Board for the year 1955.
The assessors of the various municipalities of Middlesex County completed the assessments of the local real property for 1955 and filed the assessments lists with the county [445]*445tax board as required by N. J. S. A. 54:4-35. The county board, as the agency charged with the duty of supervision and control over the assessors (N. J. 8. A. 54:3 — 16) is directed to:
“[A)nnually ascertain and determine, according to its best knowledge and information, the general ratio or percentage of full value at which the real property of each taxing district is assessed according to the tax lists laid before the board. It shall prepare an equalization table showing the assessed valuation of the real property in each district, the ratio or percentage, if any, by which the assessed valuation should be increased or decreased in order to correspond to true value, and the true value of the real property within the district as determined by it.” N. J. S. A. 54:3-17.
Iii the discharge of this statutory obligation the county board found that the 1955 ratios of assessments to true value in the municipalities which are prosecuting the appeal, were as follows:
Carteret 14.2%
Sayreville . 17.0%
W oodbridge 14.9%
An arbitrary uniform level of 22% for county-wide equalization purposes was determined upon. It represents the median percentage between the lowest and the highest municipal ratios, namely, 13.6% in Piscataway Township and 30.9% in the City oí New Brunswick. Appellants’ assessment ratios (as well as those of all other municipalities in the county) were then projected to 100%, and in order to achieve the same relative level of true value throughout the county, 22% thereof was fixed for each as the aggregate assessments for equalization table purposes. The design of such procedure is to bring about a fair allocation of the county tax burden among the local governmental units.
Ten of the 25 municipalities appealed to the State Division of Tax Appeals. Pull hearing resulted in affirmance of the ratios determined upon by the county board. A few changes were made which are not involved specifically in the problems now presented to us. The county board had left the aggregates undisturbed in seven municipalities where [446]*446the ratio of assessments exceeded or was close to 22%. With the exception of these seven, in constructing the table for purposes of equalization the county board listed the dollar totals of the assessment aggregates at 22% of the projected true value; that is, in the true value column of the table actually 22% of that value was set forth. In its revised table in the true value column the Division listed the dollar totals of all of the municipalities at the projected 100% of true value. This alteration follows more closely the literal meaning of the statute.
Twenty-two of the county's 25 municipalities have accepted the result. Only Carteret, Sayreville and Wood-bridge bring the matter before us. They charge that the judgment of the Tax Division is arbitrary, capricious and based upon data insufficient to justify the conclusion reached as to ratios of assessments to true value.
The principles which guide and control the equalization function are thoroughly explained in the informative opinions of Justice Brennan in City of Passaic v. Passaic County Bd. of Taxation, 18 N. J. 371 (1955), and Borough of Little Ferry v. Bergen County Bd. of Taxation, 18 N. J. 400 (1955).
The statute calls upon the county board to determine “according to its best knowledge and information” the general ratio or percentage of the full value of real property in the taxing districts which is represented by the assessments, and to equalize the aggregates at full value in order equitably to distribute the tax burden among them. No specific plan or method for reaching that determination is prescribed. Any reasonable and efficient mode may be adopted. The process is legislative or guasi-legislative in character. The conclusions need not rest upon proof admissible under common-law rules of evidence, and an oversensitive regard for such rules is inconsistent with the practical and just discharge of the duty. City of Passaic v. Passaic County Bd. of Taxation, supra, at pages 386, 389.
Equalization of aggregates for the purpose under discussion is not required to be perfect. It is recognized as a [447]*447reasonable approximation, calculated to achieve the same relative valuations among the taxing districts so as to distribute the county tax burden fairly and to offset competitive undervaluation by local assessors. The aim is to minimize “so far as possible the unfair distribution of the county tax which is one result of varying average assessment ratios” among the municipalities involved. Id,, at page 381. A degree of imperfection is tolerated largely because individual assessments are not altered as the result of the table.
To facilitate the equalization procedure, it is important to develop methods of valuation which can be applied on a mass basis and which can be expected to produce a proportionate distribution of the tax burden. Id., at page 395; Murray, “Improvement in Real Estate Taxation Through Assessment-Sales Studies,” 5 Nat’l Tax J. 86 (1952).
As part of the operation, the Supreme Court declared that the board may take official notice of the Sixth Report of the Commission on State Tax Policy (1953) and:
“* * * [inlay properly consult and rely upon official records of real estate transactions and draw inferences of true values of properties from revenue stamps on recorded deeds, cf. N. J. S. A. 54:3-22, or the amounts of mortgages on such properties, and may consult appraisals in the files of mortgage lending institutions or of such agencies as the Veterans Administration or the Federal Housing Administration. And the sampling need be merely reasonable in light of the purpose and not exhaustive * * *. The boards may also, and should, take official notice and give consideration to the real estate average assessment ratios for municipalities of the county as determined by the Director of the Division of Taxation pursuant to N. J. S. A. 54:1 — 35.1 et seq. [the ‘school aid ratios’].” Borough of Little Ferry v. Bergen County Bd. of Taxation, supra, 18 N. J., at page 405. (Emphasis ours)
The only condition upon the use of such information is that the municipalities he informed that it is being consulted and considered and be given an opportunity to meet it. Little Ferry case, supra, at page 405.
When the equalization table has been issued and is brought before the Division of Tax Appeals by a municipality, its function is the same as that of the county board. Upon a showing “merely of error by the county board as [448]*448to the aggregate of any one of the taxing districts” the state body must undertake the task of revising and correcting the table. City of Passaic v. Passaic County Bd. of Taxation, supra, 18 N. J., at page 393.
However, after review of the table, the action of the Division is presumed to be correct and the courts will not interfere in the absence of an affirmative showing that it was arbitrary, capricious or unreasonable. State, Weehawken Twp., Pros. v. Roe, 36 N. J. L. 86 (Sup. Ct. 1872); see Carls v. Civil Service Commission, 17 N. J. 215 (1955).
Appraisal of the reasonableness of the decision of the Division must be engaged in with the realization that “[t]rue parity of the aggregates of the several municipalities cannot really be achieved without elimination of the inequities among individual assessments within all municipalities of the county.” Passaic case, supra, 18 N. J., at page 379.
It is a matter of universal knowledge that equalization between individual assessments within the taxing district is the heart of the real estate taxation problem. Murray, “Improvement in Real Estate Taxation,” supra, at page 86. “Average assessment ratios are at best only estimates of the assessment experiences for individual properties * * *. Even the most accurate average suggests a uniformity of experience within taxing districts which does not exist * * *. These variations are serious and widespread. * * *” Sixth Report, supra, quoted in the Passaic case, supra, 18 N. J., at page 380. Exactitude can be accomplished by the uniform application of any common assessment standard, whether it be true value or some other rule. In our State the Legislature has willed that the measure shall be true value and both the assessors and the courts must act accordingly.
The only completely adequate way to equalize assessments is to make them equal in the first place. Equalization among municipalities within a county for a limited purpose is no substitute for good local assessments conformable to the legislative will. Murray, "Improvement in Real Estate Taxation,” supra, at page 87. However, it has been said [449]*449that the supervisory interest of the State and county through the medium of equalization is being reflected in an improvement in the quality of local assessments. Myers and Stout, “Recent Trends in Property Tax Equalization” 3 Natl Tax J. 179, 186 (1950); and see Lee, “Stale Equalization of Local Assessments,” 6 Id. 176 (1953).
These preliminary observations bring us to a consideration of the equalization table under attack.
At the outset of the hearing before the Division of Tax Appeals, the panel announced to counsel that notice would be taken of:
(1) The Sixth Report of the Commission on State Tax Policy;
(2) Information relating to the 1954 or 1955 reports of the Director of Taxation;
(3) Studies and information described as material in the Passaic and Little Ferry eases;
and that evidence would be received concerning the study of sales which had taken place in the county.
The proof adduced disclosed that in the early summer of 1952 at the direction of the State Tax Division the county board called together all of the assessors and informed them that an equalization table would be prepared for 1953 and that sales ratio studies were going to be utilized in order to fix a ratio of assessments to true value to be applied in the 25 municipalities in the county.
Pursuant to this plan, in June of that year the board engaged Mrs. Ethel Yahnell as a tax research analyst to make the sales studies. She was still so employed at the time of the hearing before the panel. In the intervening period equalization tables, which reflected her work, were issued for 1953 and 1954.
In July 1954 the board sent a letter to all assessors referring to the inequality in assessments within and between municipalities, and directing them to increase their assessments to at least 22% of true value. The letter reported also on the comparative ratios for the first half of 1954 as indicated by the current sales study and suggested their use as a guide for the 1955 assessments.
[450]*450The ratios set forth for the appellants were:
Carteret . . 15.2%
Sayreville . 16. %
Woodbridge 14.4%
In addition, the assessors were given photostatic copies of all abstracts of deeds showing the revenue stamps and they were instructed to consider them in verifying the figures of the board.
In the course of Mrs. Yahnell’s work, she examined 10,300 abstracts of deeds arising from real property sales in the 25 municipalities of Middlesex County in 1953. After computing the selling price of each parcel from the revenue stamps, see Int. Rev. Code of 1954 § 4361, a comparison was made with the 1954 assessment. The result was used as the basis of the 1955 ratio subject to whatever adjustment was necessary to accommodate any new or increased assessment which occurred in 1955.
The 10,300 abstracts, by a two-step process, were refined to an ultimate net of 2,692 in reaching the final computations. The diminution process was engaged in to eliminate sales of vacant land, sheriff’s sales, transfers between related persons, those involving conveyance of part interests, and those subject to mortgages where the amount of the mortgage could not be ascertained accurately. Moreover, when there had been extensive residential development in a municipality and many houses of the same type had been erected, a uniform price-assessment range would appear frequently. In these situations only one sale would be used in order to avoid giving such transactions what was regarded as an undue impact on the study. Their stripping away was designed to lay bare a hard core of sales which would have a high potential of significance when the relation between assessment and true value is being sought. The practice of eliminating unusable sales manifestly lessens the margin of possible error and is acknowledged to be an important ingredient of the procedure. Murray, “Improvement in Real Estate Taxation,” supra, at page 91; National Associa[451]*451tion of Tax Administrators, Guide for Assessment-Sales Ratio Studies, 6, 34 (1954).
To arrive at a municipal ratio, first the percentage of assessment to true value (as indicated by the revenue stamps) of each usable sale in that municipality was determined. There was no breakdown of the sales into classes of property or computation of ratios on that basis (except that in the case of a typical residential development, only one sale was used as discussed supra). Then all of the resulting ratios were added, divided by the number thereof, and the quotient represented the ratio for the municipality.
This rather simple mathematical formula was the critical factor in the creation of the equalization table by both the county board and the Division. However, the evidence shows that members of the county board had made many inspections of properties, estimated at 1,000, throughout the county in connection with tax appeals and otherwise. On such occasions they would look over the neighborhood generally and get “a fair idea of what the assessing situation was.” Moreover, they talked with local interested people about values of property; they considered the Sixth Report, to which reference has been made, and they went over with the Commissioner (sic) the findings as they related to Middlesex County. And the record provides support for the view that the county board exercised an informed judgment in deciding to accept the average municipal ratios which evolved from the described sales study.
Appellants’ most severe criticism of Mrs. Yahnell’s work is that no classification or “stratification” of the sales into types of property was made. Particular emphasis is placed upon the asserted desirability for doing so with respect to industrial property, which it is said is assessed at a higher percentage of true value than other types.
If it be true that various classes of property are assessed at varying percentages of true value — a discriminatory and illegal practice and one which we will not assume is followed in a particular municipality in the absence of adequate proof — a more accurate overall average ratio might [452]*452be derived in some cases from classification. It may be noted that the Director of the Division of Taxation, in the studies to assist him in the performance of his duties under N. J. S. A. 54:1-35.1, ei seq., uses four categories of sales. They are (1) vacant land, (2) residential, (3) farm, and (4) other. Switz v. Township of Middletown, 40 N. J. Super. 217 (App. Div. 1956). “Other” includes industrial and business properties, apartments, flats and specialties.
Of course, there may not have been sufficient responsible and significant data currently available to permit stratification. This is the Sayreville status, hereafter discussed. And unless a showing is made that its absence renders the sales sampling actually used so inaccurate as to be arbitrary and contrary to reason in the light of the broad equalization objective, the table should not be set aside.
This brings us to the specific charges against the ratios approved by the Division for appellant municipalities.
I.
CARTERET.
The county board's analyst examined 156 sales in the borough. Most of them were residential; four were industrial and commercial. They were refined through study to a net used of 12 representative ones. They showed a 14.2% ratio of assessment to value which was accepted by the board as the basis for adjustment to 22% in order to bring about equalization.
The elected assessor asserted that he assessed 14 major industries in Carteret at 40% of value. But it appeared that his valuations were taken over in 1948 from his predecessor in office and he had made very little changes in them. Any sales that took place subsequently were not reflected therein. When asked how he arrived at his gross figure of $11,440,480 as their market value, he said:
“We made a cursory examination and spoke to practically all of the superintendents, and from what information I could get from them I put on that figure.”
[453]*453It appeared also that his appraisals on light industrial and business properties were not based on any sales. Business properties, he said, were assessed at an average of 20%, but there was no general average. And when a sale occurred which varied from the 20% ratio, he made no revision in the assessment.
Residential premises when new construction were assessed at 12 to 13% of the purchase price; older residences at the same percentage of present value.
A resident real estate broker was produced. His testimony was to the effect that at the assessor’s request he had visited the heavy industrial plants, light industrial and business properties, and appraised them. Also, he made a spot check of residential premises. He inspected 875 properties over a period of 20 days.
In reaching his valuations of the industrial properties he did not follow the reproduction cost less depreciation method nor the capitalization method, nor did he consider any comparative sales. He used the “economic approach,” that is, he “took it for what it stands for, what it is worth, what I would figure it worth on the size of the building, the condition it is in, and the fair market value thereof. * * * I saw no figures, I got no information from anyone. * * *” He used the “method that I, as a realtor, would take for its value if I am asked to go and look at a thing and state what it is worth. I go and look at it, and measure it, and I think what it is worth.” He never owned any heavy type industrial properties and never bought or sold or participated in the sale of “property of this type, of any of these properties.”
This outline demonstrates that the borough testimony provides no reliable guide either for a determination of the true value of the real property ratables or of the percentage of assessment to the true value thereof. Certainly it cannot be said to provide a sound basis for establishing error sufficient to warrant invalidation of the table.
The Sixth Report showed the assessment ratio to be 12% for residential property and 41% for commercial properties. [454]*454The latter figure was based upon a sampling of two industrial and commercial buildings (the lands were not included), and considered in isolation they resulted in the 41% figure. As such, they furnish too insubstantial an index of the assessment practice for business properties generally in 1955 to require us to reject the methods employed by the county board.
The borough insists that it was improper to apply 1953 sales to the 1954 assessments and to use the results on the 1955 assessments in reaching ratios appearing in the table in question. In making the contention reference is made to N. J. S. A. 54:3-17 which calls upon the county board to ascertain and determine annually the general ratio of assessment for purposes of promulgating the equalization table. Therefore, it is said that the 1955 assessments alone offer the basis for fixing the percentage. But, as already pointed out, the statute does not prescribe any method of reaching the annual average ratio. Moreover, appellant does not demonstrate how the 1955 assessments would render erroneous the table ratio. On the contrary, Mrs. Yahnell testified that in arriving at the ultimate ratio for 1955, consideration was given to any added or increased assessments which appeared in the 1955 assessor’s tax list.
However, a mathematical error does appear in the record by undisputed facts. One of the industrial sales used by Mrs. Yahnell showed the percentage of assessment to value as 92%. But in listing the various percentages originating from the sales sampling this sale was included as 9.2% instead of 92%. Correction of the error means a change in the overall average ratio from 14.2% to 15.35%. We have not checked the computations involved but assume the figures are accurate since no one has challenged them. Nor does the record reveal just what difference the error will make in the amount of the county tax burden imposed. In any event, a remand must be ordered to rectify this mistake as to Carteret and to make the consequent adjustments as to all the other municipalities.
[455]*455Aside irom this error, appellant has not established any other adequate ground for a decision by us that the pertinent portion of the judgment of the Division is without substantial foundation.
II.
SAYEEVILLE.
Eive hundred sixty-two abstracts of 1953 sales in this community were examined. The number was refined to 124 net usable sales by the process already described. Only one of these was a sale of heavy industrial property; five were sales of commercial or business property, although at another place in the record the number was put at 12 or more. The net usable ones formed the basis for arriving at the average ratio of assessment to value of 17%.
Appellant challenges the reliability and sufficiency of the sales sampling data which resulted in the 17% ratio.
Primarily the attack is grounded upon the character of Sayreville’s ratables. They are said to be “in terms of dollar value,” 65% industrial and commercial property, 25% residential and 10% vacant land.
The assessor testified that the overall average ratio of his 1955 assessments to true value was “21% or better.” The industrial ratio claimed in the past and for 1955 was about 25 or 26%, although at one point he said: “Well I stated before, I think industry is assessed a little higher than the residential is.” So it is urged that the equalization table should not have ordered any increase in the borough’s aggregate.
The general thesis advanced in support of the charge of error is that since 65% of the assessable real property is industrial and since the evidence shows the assessment of industrial property at 25 to 26% of true value, the use of 124 sales, all of them involving residences except one industrial and five (according to Mrs. Yahnell) business or commercial, was a wholly unreliable method of arriving at a fair average ratio for all types of property.
[456]*456The one industrial sale was for a price of $2,700,000. The borough suggests that the only way of giving proper weight to it in relation to the more numerous residential transfers, is to add the dollar value of all sales and divide the result into the total of the assessments of the property involved. This weighted average percentage, it is claimed, represents more properly the relative significance of the transaction.
On the other hand, the county board evidence was to the effect that to follow such a formula would be to give undue effect to the single large transaction and to create an improper imbalance in the eventual average ratio. Cf. Guide for Assessment-Sales Ratio Studies, supra, at page 23. Instead the board took the actual percentage of assessment to sale price which was 33.2% and included it in reaching the total of the percentages for all the sales considered. Thus it played a part in producing the 17% average ratio for the municipality.
Moreover, Mrs. Yahnell said that when the six industrial and commercial sales were put in a separate category or classification, the unweighted average of the ratios for the class was 17.3%. And her opinion therefore was that, by including these individual percentages with those resulting from all the sales used instead of the class percentage of the business and industrial sales, Sayreville was given fair and equitable treatment.
In addition, there was substantial testimony to the effect that the one large industrial sale was not a usable one at all. The premises were sold to a pension trust and then leased back to the original owner. In this connection reference was made in the examination of one of the witnesses to page 7 of the Guide for Assessment-Sales Ratio Studies, supra, which says that sales between parent and subsidiary corporations do not present reliable sales prices because the consideration is frequently an incidental concern. And:
“On the same order are sales to insurance companies, universities, etc., when the property is leased back to the seller. In such cases the sales price is not ordinarily a reliable evidence of market value because the price received is tied up with the amount of the rental provided for by the lease.”
[457]*457Despite this view, the industrial sale being discussed was included as a usable one by the board. This was to the advantage of Sayreville, in comparison with its rejection as non-usable.
Evidence was offered to show that five large industrial properties, one of them being the plant sold and used in the board’s computations, constitute almost 90% of the total industrial assessments. An effort was then made to prove that the four unsold industries were assessed percentagewise as follows:
Du Pont......... 32.62%
National Lead . ... 22.4 %
Hercules...... 32.3 %
Sayre & Fisher Brick Company 24+ %
The one transferred to the pension fund, Owens Illinois Glass Company, on the basis of the sale price, showed a 38.9% for 1955. The five percentages indicate an average ratio of 30.2%.
An associate of the borough attorney testified that he had studied 869 deeds representing sales in Sayreville in 1954. They indicated an average assessment ratio of 20.55%.
The weighted average of the two classes, industrial and residential property, would be 25.37+%. The contrast between this ratio and the assessor’s testimony that it is “21% or better” is rather marked.
A further analysis of the testimony is revealing. With respect to the 869 deeds there was no check to determine if a mortgage was involved in any of the transactions. It was assumed that the revenue stamps portrayed the full value free of encumbrances. See Int. Rev. Code of 1954, ■§ 4361, which taxes only the buyer’s equity. No inquiry was made to ascertain if there were any intrafamily or other types of sales which the board classed as non-usable ones in connection with the problem under consideration. The only knowledge professed by the witness was that the properties were improved ones. Whether any of them were business or commercial, he could not say.
[458]*458The means by which the percentages of assessment to true value of the four major industrial properties were arrived at, cannot be ignored in evaluating that proof. The borough attorney obtained written or oral statements from officials of the companies involved as to the net book value of the premises. He was told that the figures furnished were those set forth in the 1955 corporation franchise tax return. The returns themselves were not examined nor was any effort engaged in to verify the values. The witness had no knowledge as to how the book values were determined.
There is frequently a vast difference between book value and market value. No evidence of original cost was adduced or of the depreciation policy followed by the four companies. Prom an accounting standpoint, a property may be depreciated to zero and yet have a very substantial market value. Actual depreciation from the standpoint of sale in the open market may be a virtual stranger to theoretical depreciation for accounting purposes.
The Division found that book value, as proved here, is an inadequate and impractical means of ascertaining a ratio between market value and assessment. We agree.
Some indication of the unscientific nature of the borough’s assessment practice is furnished by the procedure respecting additional assessments for improvements, such as new buildings. The assessor did not know how much the National Lead Company assessment had been increased since 1950 because of improvements. He could not say if improvements to the extent of 50 million dollars had been erected. He had never checked any construction contracts.
Two new warehouses, one 80 or 85 feet wide and 200 feet long, the other 100 by 200 feet, had been built for DuPont since 1950. These contracts were never reviewed to check the costs. When asked how the value of the new construction was arrived at, the witness said:
“A. By the same [method?] as I stated before with the National Lead or Hercules. I sat down with the heads of the company and talked things over and found out what buildings were torn down, and what buildings were added, and taking the difference in cost, [459]*459and then trying to hit at a very fair assessment figure.” (Insertion ours)
In this connection, a purely psychological factor cannot be overlooked. One member of the municipal governing body is in the employ of DuPont; another one works for Hercules Powder.
As already set forth, all of the municipalities in the county were informed in July 1954 that assessment at 22% of true value was expected. Each one was told also the comparative ratios which the study of sales had shown for the first half of 1954, and the abstracts on which the computations were based were made available. Sayreville’s ratio was listed at 16%. It does seem that in the long period intervening between this notification and the Division hearing, more substantial evidence of the true value of the four unsold major industrial plants could have been procured.
The record contains much discussion by the witnesses for Sayreville and the county board as to the method of computing the ratios which should have been pursued in fashioning the equalization table. It seemed to be agreed that at least three formulas are ordinarily acceptable: (1) the median, (2) the unweighted mean, and (3) the weighted mean. The county board adopted the unweighted mean and Sayreville contends that under the circumstances of the case, the weighted mean was the only proper one.
In our judgment, it is not necessary to compare the methods or to decide whether one should have been used to the exclusion of the other. The evidence submitted by the appellant to show the true value of its industrial properties is so vague, indefinite and unreliable as to be incapable of demonstrating either error in the formula used by the county board and approved by the Division, or such error in the average assessment ratio resulting therefrom as to justify the conclusion that the equalization table ought to be set aside.
Eurthermore, we conclude from an analysis of all of the proof that appellant failed to establish sufficiently that any higher uniform percentage or any higher average ratio of [460]*460assessment to true value was applied by its assessor to industrial property than to residential property. Nor can we say on the evidence that the Division erred in finding that no such difference appeared as to require separate classification of industrial property in computing assessment ratios or to make the average ratio of all the sales used unfair, unreasonable and inequitable.
The 1955 sales study, of which we have already spoken, did not present any substantial criticism of the county board’s ratios. The transactions were included in appellant’s computations indiscriminately and without regard to any of the considerations which are regarded as significant in establishing usable or non-usable sales. Such treatment manifestly cannot be given the probative force of the highly selective method employed by the county board.
We note that the Sixth Report, supra, gives the appellant’s overall average ratio as 16%, which is the same as that set forth in the county board’s letter of July 26, 1954 to all of the assessors. On the evidence now before us the 17% ratio cannot be characterized as so unreasonable and unfair as to require its rejection.
III.
WOODBRIDGE.
The analyst of the county board examined 2,235 abstracts of deeds covering 1953 sales in this community. By the process referred to earlier, that number was reduced to 391 net sales used in determining the average ratio. A major factor in the exclusion of so many was the large scale residential construction which had been going on in Woodbridge. Where there was a substantial number of sales of the same type from a development only one percentage of assessment to true value was adopted as in the case of other municipalities in the county. Vacant land sales were excluded. No classification of kinds of property was made. All of the net used sales were included in a single computation to arrive at the aggregate percentages. As in the other cases, the average [461]*461ratio was readied through dividing by the number of percentages. The result was 14.9%.
The objection to the unweighted mean method followed by the county board is renewed. Again it is suggested that properties in the sampling should be stratified by types and the overall average ratio should be obtained by averaging the percentages of assessment to true value of the classes.
However, as in the case of Sayreville, although to an even greater extent, there is insufficient proof to demonstrate that different and discriminatory standards or percentages are applied to various types of property as a regular practice by the assessors. And the evidence is likewise unequal to the task of establishing, to the degree necessary to warrant overruling the judgment of the Division, that calculation of the average ratio on the basis of any such illegal “standards” would result in a substantially different average ratio from that already fixed.
Much emphasis is placed upon a study of 51 abstracts of 1953, 1954 and 1955 sales within the township. Using the revenue stamps to indicate value, assessments averaged to about 20%.
There were 4,500 sales in the years mentioned. The 51 used in the study were selected by the assessors and theiT counsel. The principal clerk of the board of assessors who presented them and the percentages of assessment in each case, said she thought they were designed to represent a sampling of different parts of the township. She conceded that she had received the abstracts of each of the 4,500 deeds from the county clerk subsequent to recording. But she had never been asked by her board to figure the average ratio resulting from all of them. She was asked:
“Q. Do you have at least 50 sales under the 20% margin or ratio? A. I believe so.
Q. There are plenty of sales under the 20% ratio? A. Well, out of 2,000 [1953] and 2,200 [1954], yes, I assume there would be. (Insertion ours.)
Q. So that the 51 sales that you testified to do not represent the average ratio of all the 4,500 sales? A. No, sir.”
[462]*462One member of the board of assessors produced an additional list of 21 sales for 1954. He asserted that the assessments averaged about 20%, On being asked as to the system adopted in selecting them for presentation at the hearing, he said:
“I don’t know as there was any definite system. We just checked the sale to see what the ratio was and we used them.”
He “would say” that they were representative of the 4,500 sales.
After these two persons testified, Mrs. Yahnell went through the Woodbridge files and made a spot check of the 4,500 sales. She withdrew 92 of them at random. The average ratio of these turned out to be 13.9%. No effort was made to negative the effect of this rather portentious showing, although at the conclusion of her testimony one member of the panel said to counsel for Woodbridge:
“If there is anything you find improper, or if the computation is wrong, I want you to call it to our attention.”
Appellant rests heavily on the fact that the Sixth Report gives its average ratio as 17%, and more particularly upon a county board letter of October 17, 1952, with respect to equalization which recites the ratio as 21.5%.
Specifically, it is said that the 1952 figure signifies a drop in the assessment ratio of almost 7% between 1952 and 1955, in spite of the fact that there has been an increase of almost 10 million dollars in the aggregate. The argument overlooks that in 1952 the board, to use its language, had just made a “start” on the equalization process. So it seems reasonable to believe that the sales data at the time were more limited than when the present study was made. In addition, it appears that almost 58 million dollars of permits for taxable construction were issued between 1950 and 1954. Moreover, one member of the board of assessors conceded that they had not yet gotten around to equalizing on residential properties. The secretary of the county board said [463]*463that the value of the new homes had increased at least $1,000 each between 1954 and 1955, with no corresponding advance in assessments.
The county board in July 1954 called to the attention of Woodbridge that 011 the basis of the sales analysis the ratio for the first half of 1954 was 14.4%. The preliminary equalization table was sent to each municipality prior to a meeting thereon on January 25, 1955. Its representatives attended the meeting at which everyone was advised that all the board’s records, including the abstracts of sale, were available for perusal. They did not take advantage of the offer.
It is evident from all these facts that appellant was on ample notice as to the probable outcome of the equalization endeavor of the board. Yet it failed to provide adequate facts through proper study of the 4,500 sales of which it had records, or through appraisals of a fair sampling of residential, commercial and industrial properties. And certainly the proof that was offered as to the 51 and then the additional 22 sales, was not conducive to the conviction that they indicated a general average ratio of assessments to true value. In short, in our judgment no sufficient error in the county board table was proved to permit this court to say that the result of the grosi-legislative task should be voided.
Two further objections warrant some mention. First, it is argued that the revenue stamps on the deeds should not have been accepted as proof of value, and second, that the board committed error in refusing to inspect the property in ihe township.
It is true that N. J. 8. A. 54:3-15 provides that the board shall view and inspect “so far as possible * * * and make their revision and correction after * * * inspection.” Assuming that this comprehends the equalization process, in our view it is a directory and not a mandatory requirement. Lack of compliance or inability to do so through pressure of work should not render the equalization table invalid unless some prejudice emanates therefrom. According to the testimony, the members of the board had inspected [464]*464about 800 properties throughout the county over a four-year period prior to the hearing, in connection with appeals from assessments. It appears also that during 1954 they made approximately 1,000 visitations to properties throughout the county. During these inspections and visitations, they would look around the neighborhood and get a “fair idea of what the assessing situation was.” And in the case of Wood-bridge, 19 photographs of the business area were submitted at this hearing. Under the circumstances, absence of inspection during the pendency of the proceeding before the Division presents no ground for reversal.
With respect to the argument that the revenue stamps should not be taken as proof of value, it cannot be disputed that the sales price of property is evidence of market value, although in proceedings such as tax appeals it is not conclusive evidence thereof. City Holding Co. v. State Board of Tax Appeals, 127 N. J. L. 168, 169 (Sup. Ct. 1941). And the revenue stamps are prima facie evidence of the sales price. N. J. S. A. 54:3-22.
The stamps are not ordinarily conclusive evidence of true value because of conditions, circumstances and exigencies unrelated to such value which sometimes produce a sale. In the present situation, Mrs. Yahnell exercised great care in excluding from the usable sales for ratio ascertainment purposes those which showed even suspicion of any circumstances extraneous to those which motivate a willing buyer, not required to buy, and a willing seller, not required to sell, in consummating the transaction. And she eliminated also all eases where mortgage considerations cast doubt on the probative significance of the stamps as to sales price. Thus it is plain that the stamps on the deeds which were finally used in reaching the average ratios, were strongly indicative of a normal sale.
It may be said also in passing that in the performance of the gwasi-legislative equalization function, which of itself does not affect changes in individual property assessments, the evidentiary weight of the stamps may be more liberally recognized than in ordinary judicial proceedings.
[465]*465Here again, after examining all of the objections, some of which do not seem to require specific discussion, we cannot say that the Division acted arbitrarily and unreasonably either in approving the sales study method adopted by the county board to reach the average ratio of Woodbridge assessments or in accepting the ratio itself which resulted therefrom.
GENERAL CONCLUSION.
The purpose of equalization is to minimize so far as possible unfair distribirtion of the county tax burden, in spite of divergences in the original assessment practices. It is designed to avoid the debased values which are produced through competitive under-assessment or for other reasons. The mechanism by which the equalization table is created is not yet an exact science. The Supreme Court has said that the process does not demand precision in valuation. Borough of Little Ferry v. Bergen County Bd. of Taxation, supra, 18 N. J., at page 404, 405. Occasionally minor inequities may flow therefrom, but they cannot be permitted to cancel the entire overall adjustment. These imperfections will undoubtedly be brought to an irreducible minimum as experience mounts and sales and other data accumulate.
In his present modus operandi (as we were informed on the arguments of the various county cases now before us) the Director of Division of Taxation receives an abstract of every sale throughout the State. In turn, he sends it or a copy to the proper municipal assessor for comment. In this way the assessor is put on notice that the sale is being considered and that his report thereon will aid in the evaluation of the transaction for equalization purposes.
With the passage of time and improvement in the technique of the agencies, it is only natural to expect that judicial appraisal of the method used in accomplishing parity of aggregates will become more exacting. In the meantime the conclusion seems unavoidable that fair and adequate internal equalization within municipalities which can be relied upon by the county board is a most desirable and [466]*466essential objective. Such realization translated into conforming action will reduce external equalization to a simple formula. For the present, in passing upon the issue of whether the method followed is arbitrary, the judicial approach must be one of reasonable tolerance.
In this connection, it must be appreciated that the addition of substantial sums to municipal aggregates frequently means a comparatively small added portion of the county tax burden because of the diffusion of that burden among the tax-paying municipalities. For example, it was said by the county board’s tax analyst during her testimony that the augmentation of the Sayreville assessment aggregates meant the assumption of an additional share of about $5,000 of the county tax. The amount was not made the subject of conflicting testimony between the parties. We refer to it for illustration purposes and not as a factor of consequence in reaching our decision.
So we conclude on the basis of the entire record that the judgment of the Division is not the product of arbitrary or unreasonable action. However, for the reasons stated, the table is remanded to it for the purpose of correcting the error in computation which is described in the portion hereof dealing with the evidence relating to the Borough of Carteret. Otherwise the judgment is affirmed.