(dissenting). The reversal herein by a majority of the court is based upon the viewpoint expressed in the prevailing opinion of another case, decided herewith {People ex rel. Reynolds v. Kinnaw, ante, p. 718). It was there said that the crucial question presented was the ratio which the assessed value of real property in the city of Albany bears to its full value. The view was taken that the Special Term, which confirmed the report of a referee, failed to give due consideration *723and weight to the assessment rates fixed by the Board of Supervisors of Albany County and the State Equalization Board. Several other tax cases decided herewith reveal a diversity of findings by the Special Term as to the rate of inequality, ranging from a low point of 65% to a high point of 72.4%. It was stipulated in all of the cases that the Board of Supervisors fixed the equalization rate of the city of Albany at 96% of true value, and there is proof that the State Equalization Board fixed the rate of assessment at 93%.
In all of the cases proof was taken under the so-called sampling method. Each side selected twelve parcels for comparison purposes and out of the twenty-four thus selected the referee selected twelve concerning which proof of value was taken. The value as thus found was compared with the assessed valuation of the parcels selected and an average rate of assessment was arrived at as a matter of mathematical computation. In each case evidence given by a member of the committee of equalization of the Board of Supervisors of Albany County was stipulated in the record, and the certificate of his committee fixing the rate of assessment at 96% was received in evidence. Apparently the referee and the Special Term, in each case rejected these official rates in favor of results arrived at by the sampling method. It is this policy that has been condemned.
While the sampling method of determining inequality has obvious weaknesses, it represents a practice long established in cases of this character, extending back at least as far as 1888, and it has at least implied approval by statute (People ex rel. Warren v. Carter, 47 Hun 446, mod. 109 N. Y. 576; Tax Law, § 293). Its- chief weakness rests upon the limited number of parcels chosen for comparison, but some limitation in that regard is necessary for the purposes of trial or otherwise a trial might be dragged to intolerable lengths. It must be freely conceded that the results of such a method are bound to vary from case to case, in view of the limited number of parcels considered, unless the same parcels are considered and the same experts sworn before the same referee in each and every case. However, it is a method, not only permitted by statute, but one which has been used by the courts in countless cases, and ought not to be lightly discarded for the certificate of some official body, at least in the absence of a legislative directive. Moreover, in the last analysis, no satisfactory method of arriving at a general rate can be devised that does, not take into consideration some form of a sampling method. The testimony *724of any expert on the subject is persuasive only to the extent that it is based on observation and comparison.
In the majority opinion referred to, considerable stress is laid upon the findings of the equalization committee of the Board of Supervisors and the State Equalization Board. But, when the proof in that regard is examined carefully, it will be found to amount to little more than an actual statement of the percentages. Thus, a member of the committee of equalization of the Board of Supervisors testified that his committee followed the percentage figures of previous years and also considered the fact that the State Board of Equalization had fixed a rate of 93%. As a matter of substance that practically covers his testimony. Nowhere is there any proof as to how the equalization committee of the Board of Supervisors or the State Equalization Board arrived at the rates they found, except as I have indicated. In the absence of such proof, the trier of the facts was not bound to accept their bare findings as against actual proof adduced under the sampling method.
Between the sampling method and the acceptance of rates fixed by the equalization committee of the Board of Supervisors or the State Equalization Board, there is no middle ground. Without doubt the certificate of either is admissible in evidence, and if there was no other proof it would suffice, but its weight is an entirely different matter where there is nothing behind it to indicate how the rate therein was fixed, and there is other proof to the contrary. No trier of the facts can analyze or weigh the naked finding of any committee or board when there is nothing to indicate on what basis the finding is arrived at. Perforce he must accept such an administrative finding on its face or reject it. If he accepts it on its face, then he must reject the statutory sampling method unless the results in both cases happen to coincide, which is most unlikely. On the other hand, if he rejects the rate fixed by the committee or the board, then there is nothing left but the sampling method, except, of course, actual sales for the purpose of comparison may be proven. But that, in itself, is merely another sampling method.
As the law now stands the issue of inequality is a question of fact and it is difficult to see how it can ever be anything else so long as tax certiorari cases aye permitted. But it cannot be handled as a question of fact if a referee is bound to accept an assessment rate fixed by some governmental body. If he is so bound he loses in fact his own independent judgment on the issue. The Legislature may, I assume, make a rate finding by a governmental agency prima facie proof which would have to *725be rebutted, but it could hardly make it conclusive and at the same time give the taxpayer the benefit of a real trial on the issue. Irrespective of this observation, I am assuming, for the purpose of this case, that a presumption of correctness follows the rate established by any governmental body, but in my view this presumption disappears in the face of actual proof taken under,the sampling method, or at least the trier of facts maji so determine.
It should also be observed that if a governmental rate is to be accepted in one case it should, in all logic, be accepted in all other taxes involving the same tax year and district. In view of the manner in which the law has been developed in tax cases, a great deal of confusion is bound to result if any such rule is applied without legislative sanction.
In all of the cases decided herewith, with the exception of tne case of People ex rel. Reynolds v. Kinnaw (supra), overvaluation in a substantial degree has been found. In fact, in all such cases the city made no attempt to justify the assessments as made. The relators should not be required to try their cases over again on the issue of inequality where that issue has been tried in accordance with long-established custom and practice. I dissent, therefore, and vote to affirm the order or orders appealed from.
Heffernan, Brewster, Bergan and Coon, JJ., concur with Memorandum by the Court: Foster, P. J., dissents in an. opinion.
Orders reversed, on the law and facts, and a new hearing ordered before an official referee to be designated by this court, with costs to abide the event. The referee is directed to report to this court solely on the question of inequality.
The court hereby disapproves all findings on the issue of inequality.