This is a local property tax case involving the assessment of a 148,860 square foot,, one-story, light-industrial and warehouse complex situated on 11.031 acres of land located at 2400 Vaux Hall Road in Union, New Jersey, for the tax years 1980 and 1981. The property is officially designated as Block 39-12, Lot 1, on the tax map of the municipality.
Plaintiff alleges that its assessment for each tax year exceeds the fair market value of the property and further, that it is being assessed at a higher percentage of true value than are other properties in the taxing district. The assessment of the subject property for each of the tax years in question was:
Land
Improvements
Total
$ 984,400
1,340,500
$2,324,900
*3Since the assessed valuation exceeded $750,000, taxpayer sought direct review of the assessment in this court pursuant to N.J. S.A. 54:3-21.
The subject of this controversy is an owner-occupied, single-story and part-basement industrial plant and warehouse complex with two small yard structures and various site improvements. The building was constructed in stages during the period 1950 to 1969. The industrial plant portion, which was built in 1950, is utilized for the manufacture of painters’ tools and consists of 49,240 square feet of manufacturing space, 16,000 square feet of office space, 6,850 square feet of warehouse and 3,510 square feet of basement area. This portion of the complex is glazed brick on concrete block construction.
The second portion of the complex, constructed in 1966, identified as a “Butler” type building, is constructed of metal-clad walls and contains an area of 49,590 square feet utilized for the warehousing of raw materials.
The third section, constructed in 1969 of concrete block, contains an area of 21,570 square feet utilized by plaintiff as a finished goods warehouse. There is also a detached unfinished boiler house containing an area of 2,100 square feet which houses heating and other equipment.
Lastly, there is approximately 162,000 square feet of paved parking and loading areas with cyclone-type fencing, wrought iron fencing in the front portion of the building and two small steel frame and metal structures which contain an area of 1,120 square feet.
Overall, the improvements consist of approximately 149,9801 square feet and are in a generally good condition. The land, measuring 11.031 acres, is generally level and primarily rectangular in shape. There is frontage on three paved streets and all the usual utilities are present.
*4 Valuation
Each party relied on one real property valuation expert. Both experts were of the opinion that the highest and best use of the subject property would be a continuation of the present industrial and warehouse utilization. Each expert estimated a single value for both years. Plaintiff’s expert was of the opinion that the value of the property for each year as of the pivotal assessment dates was $2,400,000, while defendant’s expert ascribed a value of $2,900,000 for the subject.
Plaintiff’s expert utilized three sales of unimproved land to establish a land value for the subject of $100,000 an acre, thereby giving rise to a total land value of $1,103,000 for each of the tax years. Defendant conceded that this estimate presented an accurate land value. Therefore, I will accept such land value for each tax year.2
Both experts eschewed the cost approach to value primarily because of the difficulty in measuring physical deterioration and obsolescence, both functional and economic.
Each expert utilized both the market data and income approaches to value. However, plaintiff’s expert gave greater weight to the market approach while defendant’s expert felt that the income approach should predominate. Because property such as the subject is primarily purchased by owner-users rather than investment-oriented purchasers, I am inclined to believe that the market data approach should preponderate in the reconciliation of values achieved by the two methods. However, this conclusion depends on which approach is best supported by factual data. In this case I find, for reasons to be hereinafter expressed, that the value derived by means of a market data approach will control because of a deficiency of factual supportive data in the income analysis of each expert.
In his market data approach, plaintiff’s expert relied on four sales of allegedly comparable properties. Three of the compara*5bles were in Union while the fourth was in Cranford. Plaintiff’s expert indicated that the Cranford sale had less significance than the other three because of a different tax structure and a substantially different land-to-building ratio.
The three sales in Union produced a unit price range from a low of $11.21 to a high of $15.04 a square foot.3 Plaintiff’s expert made a whole property comparison4 and ascribed a unit value of $16 a square foot for the subject, which produced a total value of $2,400,000 (rounded). While this court much prefers a market analysis in terms of specific items and their individual effect on value measured by dollar or percentage adjustments, it is felt that the whole property comparison technique as utilized by plaintiff’s expert does overcome the presumption of correctness attached to the present assessment. Riverview Gardens v. North Arlington, 9 N.J. 167, 174-175, 87 A.2d 425 (1952). This court must then appraise the testimony and determine the true value of the subject property and its appropriate assessment. Samuel Hird & Sons, Inc. v. Garfield, 87 N.J.Super. 65, 74-75, 208 A.2d 153 (App.Div.1965).
Defendant’s expert also utilized a whole property comparison in his market data approach. He relied on two sales offerings and four sales of allegedly comparable property. It should be noted that he submitted an additional sale in his report but in his testimony he asked that it not be considered because he had not inspected the property. While there was no objection by plaintiff as to the use of sales offerings, it is clear that very little, if any, probative weight can be ascribed to them. Olin Mathieson Chem. Corp. v. Paulsboro, 3 N.J.Tax 255, 262-263 *6(Tax Ct.1981). The range of unit prices, in the four sales relied on by defendant’s expert, was from a low of $13.11 to an upper limit of $21 a square foot. He concluded a value of $18 a square foot for the subject, which produced a total market approach value of $2,630,646 to which he added $277,400 of alleged excess land to arrive at a value of $2,908,046 which he rounded to $2,900,000.
I find after reviewing the record in this matter that plaintiff’s expert presented a sounder market approach and a more reasonable view as to the value of plaintiff’s property for assessment purposes. Plaintiff’s expert presented an analysis of the sales data which weighed heavier in the mind of this trier of fact than the analysis presented by defendant’s expert. Therefore, I have adopted the value conclusion of plaintiff’s expert and find the value of the subject property to be $2,400,000 for both tax years. I base this determination on the following observations.
One sale relied on by defendant’s expert was shown to be nonusable because of the conditions of the sale. This sale of 450 Clermont Terrace occurred on December 1, 1981 and carried a purchase price of $6,199,000. Defendant’s expert candidly admitted that he could not obtain any information about the conditions of the sale, yet he concluded that it represented a bona fide sale of real estate. Plaintiff’s expert, on the other hand, demonstrated that this sale represented a sale of not only real estate but of personalty, machinery, equipment and goodwill. In other words, it was a sale of a going concern and would not accurately reflect a unit price for only real estate. Plaintiff’s expert also utilized a sale of the same property which took place on July 2,1980 (17 months earlier), at which time the price was $3,300,000. Defendant’s expert conceded that this was an arms-length transaction but that the price was below market because the owner was anxious to sell. There was no support offered, however, for this conclusion. The sale at $3,300,000 on July 2, 1980 and the resale on December 1, 1981 for $6,199,000 should have indicated that there was clearly something peculiar about one of the sales. I find that the testimony of plaintiff’s *7expert demonstrated that the first sale portrayed a better indication of value than the subsequent sale.
Defendant’s expert expressed the opinion that 450 Clermont Terrace was quite similar to the subject. The December 1,1981 sale he relied on reflected a value of $21 a square foot which he adjusted to $18 a square foot for the subject. Therefore, he had to conclude that the comparable property had more value on a unit basis than the subject. If this were so and the- sale in July 1980, which reflected a unit price of $11.21 a square foot, were utilized, there would then be an indication based on defendant’s expert’s testimony that the subject would be worth less than $11.21 a square foot, a result clearly contrary to the proofs adduced.
Sale transactions should not be used as value source data unless there has been inquiry into the circumstances affecting the purchase price. Circumstances surrounding and affecting a sale must be known, especially if extraordinary terms or conditions of sale appear likely. While not controlling, I also note that the Director of the Division of Taxation concluded that the December 1, 1981 sale was not usable for the Director’s sales ratio study because it included personalty, machinery and equipment.
Defendant’s expert expressed the view that his appraisal was influenced to some extent by this sale and that if it were excluded from consideration his unit value would be $17 a square foot for the subject rather than $18 which he originally ascribed for the subject. This noticeably narrows the difference in value judgment between the two experts (plaintiff at $16, defendant at $17 a square foot).
Defendant’s expert utilized a sale of the property at 655 Rahway Avenue, Union, as a comparable. This property contained 53,900 square feet used solely as warehouse space and had a land-to-building ratio substantially lower than the subject. This sale reflected a unit price of $19.24 a square foot. The differences in size and land to building ratio between the comparable and the subject property were substantial enough to *8cause little weight to be assigned to this sale. Even defendant’s expert conceded that size of industrial plants is extremely significant when making comparisons. The sale property is approximately one-third the size of the subject. The industrial market reflects that a greater unit price is paid for smaller buildings. As the size of an industrial property increases, unit price decreases.
Additionally, on cross-examination defendant’s expert was somewhat vague with regard to some of the data applicable to this allegedly comparable property relative to ceiling heights and office space.
The remaining two sales relied upon by defendant’s expert at 410 Clermont Terrace, Union, and the north end of Wainwright Street, Union, were also relied upon by plaintiff’s expert. Of all the sales presented by both experts, these sales appear to be the most comparable to the subject. These sales reflect unit prices of $15.04 and $13 a square foot, respectively.
Plaintiff’s expert presented a more credible and persuasive analysis of these two sales to support his indicated value of $16 a square foot for the subject, as contrasted to the $18 a square foot postulated by defendant’s expert.
With regard to the Wainwright Street property, defendant’s expert did not inspect the interior of the plant. Further, he testified that it had no paved parking area, which proved to be incorrect. He indicated, on cross-examination, that the fact the comparable did have a paved parking area would have affected his value indication.
Defendant’s expert also was in error with regard to the size of the land in the comparable at 410 Clermont Terrace which he reflected at 5.24 acres while the deed of transfer indicated 6.19 acres. This error would affect the land-to-building ratio of the comparable and the ultimate value indication to be derived from this sale.
In addition to ascribing a value of $18 a square foot for the subject, defendant’s expert advanced the proposition that the subject had 2.92 acres of excess land which he valued separately *9and then added to his market data approach value. He based his excess land concept on the theory that the buildings, which measured 3.355 acres, required only 8.38 acres of land to serve the existing improvements. Therefore, 2.92 acres (11.3 acres minus 8.38 acres) was excess and should be valued separately. This procedure would have validity if sale comparisons were made with properties that did not have excess land. In this case the two most comparable of the submitted sales both had excess land and such excess would be reflected in the unit prices obtained in those sales. The land-to-building ratio for the sale at 410 Clermont Terrace was 3.69 to 1, which reflects a greater amount of excess land than the subject, while the land-to-building ratio for the sale at the north end of Wainwright Street was 2.9 to 1, also reflecting excess land, not substantially different from the subject. See American Institute of Real Estate Appraisers, The Appraisal of Real Estate, supra at 282-283. Plaintiff’s appraisal expert considered the fact that these comparables had excess land and that the value of such excess would be reflected in the unit sales prices. Further, he indicated that he adjusted his comparables if they failed to reveal excess land. For the reasons indicated I have rejected the hypothesis of defendant’s expert as to excess land.
As I indicated at the outset of this discussion both experts utilized an economic analysis. However, I find that I need not make a determination based on this methodology because neither expert supplied sufficient supportive market data for the major component in this approach, the capitalization rate. Defendant’s expert simply indicated that 9.5% was an appropriate rate of return but he could provide no documentation or support for this figure.
Plaintiff’s expert presented an overall capitalization rate based on a band of investment concept. There was no support for the equity portion. He said that the equity portion was derived from alternate investments but the record is devoid of any such alternative investments or rates. The mortgage portion was premised on the ability of the purchaser of the subject property to secure a mortgage loan at 11% for a 20-year term. *10There was no specific indication that such financing was available for the subject property on the pivotal assessment dates nor was there any market data justifying a 20-year term.
Plaintiff’s expert also used a building residual technique and supplied a wholly unsupported rate of return of 10.5%. This had no more meaning than defendant’s expert’s use of 9.5%. Absent adequate factual support for the chosen capitalization rate, this court need give little weight to the supposed value derived from a capitalization process. Skytop Gardens, Inc. v. Sayreville, 3 N.J. Tax 187, 196-197 (Tax Ct.1981); see Brick Associates v. Brick Tp., 4 N.J.Tax 510 (Tax Ct.1982) for a discussion of those cases wherein experts were admonished to support their opinions with appropriate market data.
As previously indicated, I find and determine, based essentially on the market data approach to value, that the value of the subject property for each of the assessment dates is $2,400,000.
Discrimination
In addition to the issue of the true value of the subject property, plaintiff has alleged that it was discriminated against with respect to its assessments in both 1980 and 1981 because they were fixed at a higher percentage of true value than the common level of other assessments in the taxing district for those years.
Pursuant to chapter 123 (N.J.S.A. 54:2-40.4) plaintiff is entitled to discrimination relief for both tax years. Based on my determination of true value, the subject property is being assessed at a ratio of assessment to true value of 97%. The ratio prescribed by chapter 123 was 83% for 1980 and 71% for 1981. The upper limit of the common level range, as defined in N.J.S.A. 54:l-35a, was 96% for 1980 and 82% for 1981. Since the ratio of assessment to true value of the subject exceeds the upper limit, taxpayer is entitled to have this court apply the chapter 123 ratio for each year to the true value of the subject property as found by this court.
*11However, plaintiff contends that the remedy afforded by an application of chapter 123 will not suffice because it is both inadequate and inequitable. In substitution thereof plaintiff seeks to have this court apply two unweighted, unclassified ratios (each based on a single year’s sales data) to the value of the property because plaintiff insists that these ratios reflect the common level of assessment for each tax year. Plaintiff did not offer any proof to demonstrate that there was error in the calculation of the chapter 123 ratios.5
In support of its position plaintiff offered a sales ratio study of the statistical data utilized by the Director of the Division of Taxation. The analysis consisted of 458 usable sales for 1980 and 410 sales for 1981. The record indicates that there were approximately 16,377 assessable properties or line items in Union for each year. Hence, the sales for 1980 represented 2.8% of the total line items and 2.5% for 1981. Plaintiff’s appraiser indicated that the unweighted, unclassified ratios derived from the sales data were 64.68% as of October 1, 1980 and 61.23% as of October 1, 1981.
The general coefficients of deviation which measure the deviation from the unweighted, unclassified ratio within a taxing district were:
1980 - 9.35
1981 - 10.75
Utilizing these data, along with a demonstration of ratio clusters, plaintiff’s appraiser concluded that a common level did exist and it was reflected in the unweighted, unclassified ratio for each year.
Defendant responds to plaintiff’s contention by conceding that there is an apparent difference between the ratios prescribed by *12chapter 123 and the unweighted, unclassified ratios advanced by plaintiff. However, defendant asserts that that showing alone does not in and of itself justify using a ratio other than the one prescribed by the Legislature in chapter 123. I agree.
This court clearly indicated in Murnick v. Asbury Park, 2 N.J.Tax 168 (Tax Ct.1981), that where adequate relief can be afforded by means of an application of chapter 123 it would defer to the legislative provision rather than adopt another ratio. It was understood that neither ratio offered provided mathematical perfection.
In 525 Hold. Co. v. Hasbrouck Heights, 3 N.J.Tax 206 (Tax Ct.1981), Judge Evers was confronted with substantially the same issue as here. The tax year in question was 1978. Taxpayer, in that case, argued that the unweighted, unclassified ratio as of October 1, 1978 represented the common level in the taxing district for 1978. Therefore, it contended that the unweighted, unclassified ratio should be used as the measure of relief to alleviate the alleged inequality in assessment, instead of the relief afforded by chapter 123. The court chose not to use the unweighted, unclassified ratio but rather deferred to the legislative remedy allowed under chapter 123 on the basis, primarily, of stability of assessments.
In Rudd v. Cranford, 4 N.J.Tax 236 (Tax Ct.1982), Judge Crabtree of this court found that the “evident legislative design” of chapter 123 “was to promote stability in assessments” because the ratio promulgated pursuant to chapter 123 reflects sales over a period of years. See Gaynes v. Edison, supra.
Each of the Tax Court judges in Murnick, 525 Hold. Co. and Rudd recited the principle that mathematical precision or perfection is not attainable nor is it required in the assessment of real property. As Chief Justice Weintraub indicated in Siegal v. Newark, 38 N.J. 57, 61, 183 A.2d 21 (1962), perfect relief is not possible in local property assessments. All that can be required and expected is practical justice. Does utilization of the ratio prescribed by chapter 123 for each of the years in question afford practical relief? I believe that it does in the context of this proceeding.
*13An analysis of the sales data for the fiscal year ended June 30, 1980 indicates that there were 439 sales of one to four-family houses, 11 sales of vacant land and eight sales of commercial property.6 For the year ended June 30, 1981 there were 390 sales of one to four-family houses, six sales of vacant land and 14 sales of commercial property. Although the general coefficient of deviation was 9.35 for 1980 and 10.75 for 1981, the coefficients of the individual classes range from 15.17 to 36.81. These latter coefficients, however, are of little assistance because they are derived from relatively few sales.
From the sales data presented I can conclude that most of the one to four-family house assessment-sale ratios fall within a range of 50%-69% for both tax years but I cannot conclude that there is any reasonable percentage range within which the assessment-sale ratios of other types of property (as is the subject) fall. Clearly, these data do not reveal any mathematical precision as to any specific ratio.
Moreover, a comparison of the various ratios as of the same periods does not reveal any significant differences between the chapter 123 ratios and the unweighted unclassified ratios.7 For instance, the chapter 123 ratio for 1980, promulgated on October 1, 1979, was 83% while the unweighted, unclassified ratio as of October 1, 1979 was 73.83%. The chapter 123 ratio for 1981, promulgated on October 1,1980, was 71% while the unweighted, *14unclassified ratio as of October 1, 1980 was 64.68%. I find that these differences are not substantial.
Confronted with a somewhat similar insubstantial difference between the ratios provided by chapter 123 and the unweighted, unclassified ratios, Judge Lasser of this court concluded that chapter 123 should control. Jefferson House Investment Co. v. Chatham, 4 N.J.Tax 669 (Tax Ct.1982). In so doing the court stated:
My review of the evidence leads me to conclude that the use of a ratio based on sales data for a single year is not justified. Stability of assessment requires that more than one year’s data be used. The Senate, County and Municipal Government Committee Statement attached to the 1979 amendment to Chapter 123, Assembly No. 1492&emdash;L. 1979, c. 51, which changed the ratio from a one-year unweighted unclassified ratio to a weighted classified multi-year averaged ratio, states: ‘This averaging method is designed to avoid abrupt changes in ratio from year to year and to avoid undue influence of inadequate sales samples in any single year.’
Mathematical perfection in assessment is unobtainable. Similarly, perfect relief from discrimination is also inherently impossible. Siegal v. Newark, 38 N.J. 57, 61 [183 A.2d 21] (1962). The Chapter 123 formula provides practical justice by use of multi-year sales data and a range of permitted variation rather than an exact ratio to test whether discrimination exists.
Taxpayer has not overcome the strong presumption in favor of the Chapter 123 ratio by advancing sufficient proof that the Chapter 123 ratio should be replaced with another ratio.... Neither [the] average ratios nor the ratio ranges are sufficiently divergent from the Chapter 123 ratio and common level range to lead me to conclude that the test for discriminatory assessment established by the Legislature should be abandoned in their favor, [at 683]
I see no reason to deviate from that conclusion in this case.
Relief shall be limited to an application of the ratios prescribed by chapter 123 to the true value determined by this court. This affords practical justice. Siegal v. Newark, supra at 61, 183 A.2d 21.
Judgment will be entered reflecting the following rounded assessments.
1980
Land
Improvements
Total
$ 915,500
1,076,500
$1,992,000
*151981
Land $ 783,100
Improvements 920,900
Total $1,704,000
Although the industrial and warehouse complex comprises a total area of 148,860 square feet, there is also 1,120 square feet of metal structures which, when added together, give rise to an overall improvement area of 149,980 square feet.
A land value determination is necessary solely for the administrative purpose of separating land and improvement assessment.
These units of comparison are for land and building merged.
“Whole property comparison involves a single lump sum adjustment based on a narration indicating why and to what extent the property being appraised is considered better or poorer than a comparable sale property. The estimated effect on value is presented as a single figure with no allocation in terms of the specific effect of particular factors contributing to the total.” American Institute of Real Estate Appraisers, The Appraisal of Real Estate (7 ed. 1978), at 286.
For an explanation of the methodology of the Director of the Division of Taxation in arriving at weighted and unweighted ratios and their use in local property tax assessment, see Gaynes v. Edison Tp., 2 N.J.Tax 500, 179 N.J.Super. 373, 432 A.2d 127 (App.Div.1980). Cf. Willingboro Tp. v. Burlington Cty. Bd. of Taxation, 62 N.J. 203, 300 A.2d 129 (1973), and Bayonne v. Tax Appeals Div., 49 N.J.Super. 230, 139 A.2d 424 (App.Div.1958).
For purposes of this opinion the phrase “commercial property” includes industrial property and multi-family dwellings of five or more apartment units.
Taxpayer sought the use of unweighted, unclassified ratios as of the tax years and compared these with the chapter 123 ratios for those tax years. The difficulty with these comparisons is that the chapter 123 ratio for any tax year is not promulgated by the Director of the Division of Taxation on October 1 of the tax year but rather October 1 of the pretax year, e.g., for the tax year of 1980 the chapter 123 ratio is promulgated on October 1, 1979, hence any comparison of ratios should be as of the same time periods. Chapter 123 requires the Director to promulgate the chapter 123 ratio on April 1 of the tax year; however, this is the same ratio promulgated by the Director on October 1 of the pretax year for school aid purposes in accordance with N.J.S.A. 54:1-35.1.