OPINION OF THE COURT
Gibbons, J. P.At issue is the validity of a resolution of the defendant village, made pursuant to subdivision a of section 3 of the *95Emergency Tenant Protection Act of 1974 (hereinafter ETPA; L 1974, ch 576, § 4). The resolution states that prior to a public hearing, which was held on December 5, 1978, the village “surveyed rental units for the purpose of determining the number of vacant units in each Multiple-Dwelling”, and that “as a result of the public hearing and the statistics compiled by the Village of Spring Valley relating to vacancy rates and rental conditions, the Board of Trustees finds that the vacancy rate in residential units in the Village of Spring Valley is lower than five percent”. The board of trustees of the village “resolved” that a public emergency existed “requiring the regulation of residential rents in all residential housing accommodations” in the village and that “the vacancy rate in all such housing accommodations does not exceed five percent”.
Plaintiffs in these seven declaratory judgment actions contend that the finding as to the vacancy rate was defective and that the ensuing rent guidelines, as well as the resolution, should be declared null and void. The Supreme Court, Rockland County, held in their favor. We disagree and declare the resolution valid.
The attack on the resolution is two-pronged. The first is that the survey conducted by the village of the 53 complexes containing six or more apartments (hereafter the sixes) was inadequate, so that the conclusion drawn therefrom as to the vacancy rate of the sixes was inaccurate. The other is that the failure to survey buildings containing five or fewer apartments (hereafter the under-sixes) invalidated the village’s declaration that an emergency existed, as stated by the village, “in all residential housing accommodations in the [v]illage” (emphasis supplied).
The ETPA authorizes a city, town or village to declare a housing emergency and impose local housing rent control. Subdivision a of section 3 of the ETPA (L 1974, ch 576, § 4 [§ 3, subd a]) provides, in relevant part, as follows: “The existence of public emergency requiring the regulation of residential rents for all or any class or classes of housing accommodations * * * shall be a matter for local determination within each city, town or village. Any such determination shall be made by the local legislative body of such city, town or village on the basis of the supply of housing *96accommodations within such city, town or village * * * and the need for regulating and controlling residential rents within such city, town or village. A declaration of emergency may be made as to any class of housing accommodations if the vacancy rate for the housing accommodations in such class within such municipality is not in excess of five percent and a declaration of emergency may be made as to all housing accommodations if the vacancy rate for the housing accommodations within such municipality is not in excess of five percent.”
Subdivision a of section 5 of the ETPA (L 1974, ch 576, § 4 [§ 3, subd a]) states that “[a] declaration of emergency may be made * * * as to all or any class or classes of housing accommodations in a municipality, except”, and it then lists exceptions in 11 numbered paragraphs. Among these are under-sixes; public housing; housing owned or operated by a hospital, convent, monastery, public institution, school or college; hotels and tourist homes; and motor courts.
On August 28, 1978, the then Village Attorney sent letters and questionnaires to the owners of 53 buildings containing 4,786 apartments. She had obtained the names and addresses of the owners of apartment buildings having six or more sewer units from the assessment records.1 The questionnaire requested, inter alia, the number of units and the number and identity of the vacant apartments as of September 5, 1978. The letter stated that if the Village Attorney received no answer, she would assume there were no vacancies. The Village Attorney testified that she sent no letters and made no inquiry as to the under-sixes. On cross-examination the Village Attorney was asked whether she surveyed rooming houses and she said no. Apparently, she also made no survey of other exempt classifications, such as hotels, motor courts, convents, monasteries, and school dormitories (although included in her survey of the sixes were “low income cooperatively]” which apparently were exempt pursuant to section 5 [subd *97a, par (3)] of the ETPA [L 1974, ch 576, § 4 (§ 5, subd a, par [3]), as amd L 1978, ch 655, § 137]).
At the trial the plaintiffs introduced into evidence informal and apparently incomplete handwritten notes of the inspectors who were assigned by the Village Attorney to ascertain the vacancies in the sixes whose owners had not responded to the August 28, 1978 letter. These notes reveal that 12 of the 31 sixes visited by the inspectors had no vacancies and that in the case of 7 of the visited complexes, the superintendents refused to give the requested information without the landlords’ approval. The landlords apparently failed to give such approval.
On October 11, 1978 the Village Attorney sent a follow-up letter and another copy of the questionnaire to those who had neither responded nor permitted inspection, and she warned that if there were no response by October 20, 1978, she would assume that there were no vacancies. On November 6,1978 she reported to the board of trustees that she had the requisite information as to 37 of the sixes since 18 of the 53 had responded in writing and 19 others had been inspected. Based on this survey and the assumption of no vacancies as to the 16 nonanswering, uninspected complexes, she concluded that the vacancy rate of all of the sixes on September 5,1978 was less than 2%.2 When asked, in effect, how a survey limited to sixes could be a proper basis for a resolution declaring that the vacancy rate “in all [residential] housing accommodations [in the village] does not exceed five percent”, she answered: “The [ETPA] gives us two options in declaring an emergency as to all housing accommodations or to declare an emergency as to a particular classification. Insofar as the declaration of emergency was concerned, the emergency was declared as *98to all housing accommodations, meaning all housing accommodations that could be included under the act' (emphasis added).
The assessment records revealed, as indicated by a search made by a witness produced by plaintiffs, that there were 63 properties in the village with 3 to 5 apartments or sewer units and that the total number of such apartments or units was 225.
The trial court held that the village’s acceptance of its attorney’s assumption of no vacancy in the 16 unresponsive sixes rendered the resolution invalid because “the owners of the real property were deprived of the full benefit of such ownership * * * by an assumption made rather than by an accurate and complete survey which was required by law”. Under the circumstances, we disagree.
It is to be noted that although the enabling statute (L 1974, ch 576, § 4 [ETPA, § 3, subd a]) requires, as a basis for declaration of an emergency, that the vacancy rate of any class or all of the housing accommodations be “not in excess of five percent”, no method is stated as to how this fact is to be ascertained. The statute states only the generality that it “shall be a matter for local determination within each city, town or village” (L 1974, ch 576, §4 [ETPA, § 3, subd a]).3 Although such determination may not be made on less than reasonable grounds, we see nothing in the statute requiring, as stated by the trial court, a “complete survey”, if by that term it is meant that information as to all of the buildings in the relevant classification must be obtained. “Survey” as used in the Local Emergency Housing Rent Control Act (L 1962, ch 21, § 1) could not possibly mean this, if for no other reason than the tremendous number of buildings involved. Further, assuming that a “survey” was indeed required (cf. Seasons Realty Corp. v City of Yonkers, 80 Misc 2d 601, 607), that term is defined in Webster’s New International Dictionary as “a study of a specified * * * aggregate of units *99* * * with respect to a special condition or its prevalence or with the objective of drawing conclusions about a larger * * * aggregate”. Certainly a review of 69.8% of the relevant complexes (i.e., 37 of the 53 sixes) is sufficient for “drawing conclusions about [the] larger * * * aggregate”.
We believe that a more relevant claim of defect, then, is not whether information as to the vacancies and occupancies of all of the sixes had to be obtained, but whether the village’s attribution of zero vacancies to the sixes owned by the 16 continuing nonresponders fatally tainted the conclusion that the vacancy rate of all of the sixes was less than 5%. We must consider this in light of the fact that the village’s declaration of emergency was a legislative act and therefore presumptively valid. The degree of proof required for a successful attack was formulated by this court in De Sena v Gulde (24 AD2d 165, 169 [opn by Hopkins, J.]), as follows: “When a municipal legislative body enacts an ordinance, a presumption of validity attaches to its resolution (Rodgers v. Village of Tarrytown, 302 N.Y. 115; Shepard v. Village of Skaneateales, 300 N.Y. 115). The presumption of validity has the effect of (1) imposing the burden of proof on the party questioning the ordinance; and (2) sustaining the ordinance if the propriety of its enactment is fairly debatable. The content of the burden on the assailant is sometimes said to extend further than a mere preponderance of the evidence to proof beyond a reasonable doubt (Wiggins v. Town of Somers, 4 NY2d 215; but, see, Thomas v. Town of Bedford, 29 Misc 2d 861, 866, affd. 15 AD2d 573, affd. 11 NY2d 428). Still, the presumption is not irrebuttable (Arverne Bay Constr. Co. v. Thatcher, 278 N.Y. 222), and perhaps we may best rationalize the presumption as a reminder of the force of legislative judgment which must be supported by the courts if there is ‘any state of facts either known or which could reasonably be assumed’ on which the ordinance could be based (United States v. Carolene Prods. Co., 304 U.S. 144, 154; cf. Town of Islip v. Summers Coal & Lbr. Co., 257 N.Y. 167).”
In support of our conclusion that the resolution is valid, we note the following:
1. The resolution minutes state that at the public hearing the Village Attorney said that “[s]everal apartment *100owners did not respond to the questionnaire and did not permit inspection of the apartments. In those cases, the Village considered that no vacancy existed” (emphasis supplied). It does not appear that the representative of the Rockland County Apartment Owners Association (or any of the other persons who were present) contested the assertion that the assumption of no vacancy was made only as to owners of sixes who had refused inspection after failing to respond to the original letter and questionnaire.
2. The owners of the nonresponding sixes were given a second opportunity, after their refusal to permit inspection, to respond to a newly sent questionnaire and failed to do so despite the repeated caveat that “[i]f we do not hear from you, we will have to assume that there were no vacancies * * * on September 5, 1978”.
3. The fact that inspection of the 24 complexes whose nonresponding owners permitted inspection revealed that half of them had no vacancies and that owners who might have had a low percentage of vacancies had a self-interest in not disclosing this information, permitted the reasonable inference that the owners who did not permit inspection had a very small number of vacancies or none — especially since such owners had been twice warned of the consequences of no response.
4. The fact that the “less than two percent vacancy rate” found by the Village Attorney translates into the probability of substantially less than a 3% vacancy rate for the 37 complexes as to which vacancy data had been obtained4 reasonably permitted the inference that the vacancy rate of the 53 complexes was substantially less than 5%.
5. The sending of the two sets of letters and questionnaires and the interim inspections (permitted and nonpermitted) made by five inspectors constituted a good-faith effort to obtain a reasonable survey.
6. Plaintiffs’ expert witness indicated, in response to a hypothetical question asked by plaintiffs’ counsel, that data compiled from 16 apartment complexes with approximately 1,840 units was not sufficient to establish a statistically sound vacancy rate for 53 apartment complexes with *1014,786 units. However, the record reveals that the necessary data was received as to 37 of the complexes and that 16 was the number of the complexes as to which no information was received.5
7. Since a good-faith study was made based on precise data obtained from a substantial majority of the complexes, it would be anomalous to hold that those who refused to co-operate with the statistical study should benefit from their stubborn and studied silence. If it be argued that the landlords who co-operated should not suffer the consequences of the non-co-operation of the others, the answer is that it would have been a simple matter for plaintiffs to produce at the trial herein, by subpoena if necessary, the relevant statistics of the 16 noncomplying sixes.
The presumption of validity casts the burden of proof upon the plaintiffs who are questioning the legality of the village’s declaration of public emergency (De Sena v Gulde, 24 AD2d 165, supra). This burden has not been met. Accordingly, the declaration prevails against the attack based on the village’s assignment of a zero vacancy rate to the 16 nonresponding sixes.6
The other alleged defect, as aforestated, is that there was no survey or consideration of the vacancy rate of the undersixes. Plaintiffs acknowledge that pursuant to section 5 of the ETPA (L 1974, ch 576, § 4 [§ 5], as amd L 1978, ch 655, § 137, eff July 25,1978) this class was one of the 11 kinds of housing accommodation that could not be the subject of a declaration of emergency. They, nevertheless, argue that the village’s undifferentiated finding as stated in its resolution, “that the vacancy rate in residential rental units in [the village] is lower than five percent”, and its subsequent resolution that “a public emergency exist [sic] requiring the regulation of residential rents in all residential housing accommodations” in the village, mandated an analysis *102of the under-sixes as well as the sixes. They argue that subdivision a of section 3 of the ETPA (L 1974, ch 576, § 4 [§ 3, subd a]) gives the village the option of declaring an emergency either as to all housing accommodations or “any class of housing accommodations if the vacancy rate for the housing accommodations in such class * * * is not in excess of five percent”, and that the village’s choice of the former required it to make analysis of the vacancy rate of all accommodations, including the under-sixes.
Plaintiffs’ argument presupposes that if the village had declared the emergency only as to sixes, a survey limited to that class might have passed muster. However, the fact remains that whether or not full obeisance was formalistically made to the wording of subdivision a of section 3 of the ETPA (L 1974, ch 576, § 4 [§ 3, subd a]), section 5 of the same act states that a declaration of emergency cannot be made as to under-sixes, any more than it could be made as to college dormitories or convents. As testified to by the Village Attorney at the trial, “the emergency was declared as to all housing accommodations, meaning all accommodations that could be included under the act”. Although this is the self-serving statement of the counsel who imperfectly drafted the resolution, it is, indeed, the fact that no declaration of emergency could legally attach to the under-sixes. As stated by Justice Shapiro in the case of Matter of New York City Tr. Auth. (Thom) (70 AD2d 158, 172, affd 52 NY2d 1032), “[w]e may not ignore the ‘end’ merely because the ‘means’ was expressed in an incomplete manner when the result would be an absurdity”.
Plaintiffs, nevertheless, argue that the decision of this court in Central Plains Co. v City of White Plains (48 AD2d 326) mandates a survey of all exempt housing (i.e., the 11 classes of housing accommodations as to which, per section 5 of the ETPA, a declaration of emergency may not be made), as a condition to a declaration of emergency. There the plaintiff landlords sought to nullify the imposition of rent control in the City of White Plains because the city’s survey, showing a less than 5% vacancy rate, included the input of a particularly low vacancy rate of one of the exempt classes, to wit, public housing. The issue was whether this exempt class could properly be included, not *103whether such inclusion was mandated. It was raised in the context of the fact that the parties agreed that, but for the inclusion of the exempt class of public housing, “‘that survey would have established a vacancy rate of in excess of 5% thus precluding a declaration of emergency’ ” (Central Plains Co. v City of White Plains, supra, p 329). The court held that the inclusion of public housing in the survey was appropriate. It said (p 330):
“The fact that the Act specifically precludes a local government from regulating certain enumerated housing as defined in subdivision a of section 5 simply embodies the legislative restriction that housing already regulated should not be burdened with additional local regulation. But this directive has no bearing on the total number of housing units which are in fact available in a local area. In order to determine this a municipality must, as the City of White Plains has, survey all units within its city confines. The term exempt housing means, therefore, exempt from regulation under the Act, not exempt from consideration in determining vacancies * * *
“The plaintiffs may be correct that the exempt housing is always fully occupied and therefore an emergency situation may exist at all times since the vacancy rate in the nonexempt housing would have to be extremely great to offset the zero vacancy rate in the exempt units (see Amsterdam-Manhattan Inc. v City Rent & Rehabilitation Administration, 15 NY2d 1014, 1015-1017 [dissenting opn.]). However, it should be noted that the alleged full occupancy in the exempt categories may be an indicator of the unavailability of housing in the nonexempt sector. And, as previously noted, it is the scarcity of housing in an entire community which triggers an emergency declaration for an entire city.”
Plaintiffs focus upon the phrase in the decision in the Central Plains Co. case (supra, p 330) that “a municipality must * * * survey all units within its city confines” and argue that this is an absolute. They disregard the factual background that the parties in that case stipulated that the inclusion or exclusion of the large number of public housing apartments and the latter’s proven low vacancy rate were the controlling factors as to whether the total *104vacancy rate was less or more than 5%. Here, on the other hand, the landlords submitted no proof at the public hearing or at the trial that the vacancy rate of the 225 apartments in the houses containing 3 to 5 apartments could possibly shift the balance to above 5% despite the fact that a survey of 69.8% of the sixes indicated a probable vacancy rate in the 4,786 apartments contained in all of the sixes of less than 3%.
The court in the Central Plains Co. case could not have literally meant that the units of all 11 of the exempt classes must be surveyed, since the record on appeal in that case includes a table which lists 6 of the 11 types of buildings that are excepted in section 5 of the ETPA as “Not surveyed” (underlining in original). If plaintiffs’ absolutist interpretation were correct, the failure to survey convents, asylums, motor courts and tourist homes (L 1974, ch 576, § 4 [§ 5, subd a, pars (6), (8)]) would invalidate a declaration of emergency, and the court in the Central Plains Co. case (supra) would have been required to nullify such declarations.
The issue in the Central Plains Co. case was not whether all exempt classes had to be surveyed, but whether one particularly large exempt class could be included in the survey where there was proof of a proximate relationship between such inclusion and the presence or absence of an over-all 5% vacancy rate. The decision that the inclusion in such case was valid was, inter alia, an illustration of the presumption of validity of a legislative determination of a municipality. The same presumption should be applied where a municipality chooses to exclude from its survey a comparatively small exempt class where there was no indication of a proximate relationship between its vacancy rate and the presence or absence of a vacancy rate in excess of 5%. The statement of the court in the Central Plains Co. case (48 AD2d 326, 330, supra), that “a municipality must * * * survey all units within its city confines” was dictum that the court itself did not follow. Likewise, we decline the invitation to follow it.
A commonsense approach must be applied. Plaintiffs failed to produce any proof indicating that despite the comparatively small number of under-sixes and that there *105was no general reason to assume that the vacancies in the under-sixes in Spring Valley were unusually high, a survey of the under-sixes would have tipped the balance to an over-all vacancy rate in excess of 5%. Plaintiffs have failed to meet their burden of proof.7
Accordingly, the judgment dated November 3, 1982 should be reversed, and the two judgments dated November 18, 1982 should be reversed insofar as appealed from, on the law, the resolution of the village, dated December 5, 1978, declaring an emergency requiring the regulation of residential rents should be declared to be valid and the moneys held in escrow by the Rockland County Clerk as “excess rents” should be returned to the tenants who paid them.
. The Village Attorney testified as follows: “It is conceivable that some of the letters that went out were to buildings that had five units or four units, because the numbers on the assessment rolls referred to the sewer units which are more than the actual number of apartments”. For the reasons hereafter stated in footnote 7, we believe that this did not taint the survey of the sixes.
. At trial, the appellant village’s trial counsel did not ask the Village Attorney to state the details of the survey. Indeed, the court refused to permit trial counsel for the village to elicit from her the particulars of “how she determined the two percent”. As a result, the record does not include the raw data of the number of vacancies and the total number of apartments in the 37 complexes as to which full information had been received. If the vacancy rate pursuant to the Village Attorney’s formula (based on including the input of zero vacancies from the inspection) was 2%, and if we assume that the average number of apartments and the vacancy rate of the 16 nonresponders was equal to those of the responders, then algebraically the vacancy rate of the 37 responders would be 2.86%. (This is based on the following [1] assign x as the vacancy percentage of the 37 responders; [2] based on the Village Attorney’s assignment of zero vacancies to the remaining 16 of the 53 complexes, 0 = .02; [3] therefore, x = .0286).
. We note, by comparison, that the State rent control law relating to New York City (Local Emergency Housing Rent Control Act, L1962, ch 21, § 1; amd L 1963, ch 393, § 1; L 1965, ch 318, § 1; L 1966, ch 13, § 1; L 1967, ch 657, § 1), after stating that the declaration of an emergency “shall be a matter for local determination” states that this is dependent upon the making of “a survey which the city shall cause to be made of the supply of housing accommodations” at least once every three years.
. See n 2.
. The expert further testified that a 95% response rate was necessary to obtain valid statistical data, an opinion which we deem incredible in the absence of proof that the statistics obtained from 69.8% of the complexes indicated a vacancy rate so close to 5% that a more substantial percentage of the complexes had to be surveyed.
. Pragmatically, our decision is not the last word, since, at any time after the declaration of emergency, the municipality must declare it at an end upon being shown that the vacancy rate now exceeds 5% (L 1974, ch 576, § 4 [ETPA, § 3, subd b]).
. We add the fact that, as testified by the Village Attorney, while “[i]t is conceivable” that some of the letters sent to owners of buildings having six or more sewer units might have resulted in inclusion in the survey of buildings having less than six apartments, such could not have tainted the survey. We note that plaintiff submitted no proof that such was the case. Further, since it is necessarily plaintiffs’ implicit argument that under-sixes had a higher rate of vacancies than sixes, under such hypothesis the leaching of some of the under-sixes into the survey of the sixes would have resulted in a higher vacancy rate.