OPINION
DUMBAULD, Senior District Judge.In the present posture of the case at bar,1 the task of this Court is to make findings called for by the Court of Appeals, on remand. Pittsburgh Press Club v. U. S., 579 F.2d 751 (C.A.3, 1978). The findings previously made by this Court with respect to five topics specified by the Court of Appeals (see 426 F.Supp. at 555-56) were found to have been based upon inadmissible evidence, and hence unsupported by the record and therefore “clearly erroneous.” 579 F.2d at 760. New findings regarding the same five topics seem now called for. 579 F.2d at 762. The new findings are to be “based on the record of the first trial, and so much of the record of the second trial as has not been held to be inadmissible by this court” without any additional evidence. 579 F.2d at 762. This Court is to “rule on the admissibility of and weight to be given to the first survey” (579 F.2d at 762) and also determine “net profits” in accordance with the ingenious and innovative rule elaborated by the Court of Appeals rather than in accordance with the definition of that term under generally accepted accounting principles. 579 F.2d at 760-62. Finally, in the event of a holding against the Press Club, the Court is to determine whether revocation of its exemption could properly be retroactive as of fiscal year 1967. 579 F.2d 763.
Turning first to the preliminary issue of delimiting the universe or corpus of evidence to be considered, we note that the first survey is admissible, and in fact has been admitted at the first trial, without *324objection. All evidence received at the first trial was considered at the second trial, and our present decision, as noted above, is to be based on all evidence now in the record, except that excluded by virtue of the second Court of Appeals opinion. See 579 F.2d at 762.
With respect to the weight of that survey, the significant considerations are well elucidated in note 22 of 579 F.2d at 762. The chief weakness is possibility of faulty memory. It would seem that with respect to the nature of an affair given at the club by a member, the member’s memory would be more reliable than with respect to the date of the affair, the number of persons served, the amount paid, and similar details. It would also seem that selection of the largest functions as the sample would be favorable to the IRS rather than to the club, since the evidence indicates that it is the entertainment of large numbers of people that is most likely to be an outside affair, whereas common sense as well as the experience of the IRS indicates that there is “a presumption that a genuine guest-host relationship is present if the group is eight or less” (579 F.2d at 753). Hence the data in this study are entitled to some weight, and in considering the specific issues to be adjudicated we shall utilize the contents of the study to such extent as seems appropriate under the circumstances.
It will also be proper to utilize evidence from the second trial except to the extent that it was held inadmissible by the Court of Appeals. In particular helpful financial information and figures may be gleaned from charts prepared by accountants and other data derived from the club’s books of account, and not affected by the subjective interpretations and defects invalidating the responses received in the second survey.
1. Amount of outside business
The first figure to be obtained in determining the amount of outside business is the total banquet business2 for the years in question. Figures from the club’s books show the following which we find as a fact:
Total Banquet Number of % Year_Restaurant & Bar Receipts_banquets_Banquets
1967 477,504 87,975 377 18.42
1968 537,005 145,087 3 698 27.01
1969 571,482 169,8794 832 29.72
1970 620,676 214,700 735 34.59
1971 696,521 231,834 908 33.28
Total 2,903,188 849,475 3,550
Average 580,637 169,895 710 28.60
Source: Transcript, 2d trial [hereafter, Tr. II], 20-34; Ex. C-l through C-6.
The next step in calculation will be an attempt to determine how much of the total banquet business is outside business and how much is inside business.
*325It will be noticed how closely 1969 corresponds with the average for the entire five year period. This confirms the propriety of giving weight to the first survey, which dealt with that year.
That survey (PX-12 and 13) was prepared during negotiations with the IRS and was a sample compiled to convince the appellate level that the examining agent Marcolini’s RAR report (PX-18) had overstated the outside income. The club chose that year because it was the year which showed the highest percentage of outside income in any year covered by the RAR. The club selected the 25 events showing the highest dollar amount for that year, and received 20 responses (of which only 19 appear in PX-13).
According to the club’s analysis (PX-12) the amounts clearly not outside income were $13,328 or 77.4 percent; doubtful were $1,910 or 11.1 percent; those clearly outside income were $1,989 or 11.5 percent, out of a total billing of $17,227.
Independent analysis of the data in PX-13 shows these three categories as $12,-506.25, $2282.15, and $1988.67, respectively, out of $16,777.07 total billing.5 The averages of each category were $1136.93, $571.29, and $497.17, and the average as a whole was $883. The percentages are 74.5, 13.6, and 11.8 respectively.
In accordance with a comment at the trial that perhaps those who did not respond to the questionnaire knew that they could say nothing beneficial to the club’s case, we endeavored to ascertain who were the 6 sponsors unaccounted for. We could not find in the record a list of persons to whom the questionnaire was mailed. But the record does permit an independent compilation of the 25 highest billings for the year, and comparison of the lists might reveal the names of the six who did not respond.
The list of the top 25 ($750 and over) drawn from PX-C-4, schedules T-l through T-21, shows a total billing of $30,-056.90 and only 7 of the club’s top 19 appear on this list.
The 7 common items, amounting to $9926.52, and treated as inside income in the analysis of PX-13, will be so treated in the allocation of items in the new list. The $1103.45 item of W. H. Ferguson for Duff’s Business school will be treated as outside income, like the $183.92 item for the same member in PX-13. Items of “special billing” where no member’s name is given but merely the organization (Cromler-McCormick,6 $1480.55 and Institute of Architects $799.64, will be considered outside income. Don Malinger ($1837.50 and $1077.60) was apparently promoting his own business interests, and the same is true of James Shuman ($1643.34), Walter Bigham, Jr. ($1540.50), P. J. Beta ($1111.58), and E. H. Williamson ($766.60). Likewise the event sponsored by John McGovern for Mellon Institute would appear to be inside income ($1221.04). Richard Murphy’s party for the Insurance Club of Pittsburgh is also to be treated as a business item ($1089.10).
On the other hand, an item of $942.17 for Kaufmann’s Glee Club would seem to be outside income. The lunch sponsored by Gloria Ferris ($846.48) was apparently for her own business purposes, as was a second item of M. Roberts for Sykes ($805.91) which should receive the same treatment as the earlier item in the club list ($1079.30).
With respect to James Smith’s dinner and luncheon for the Allegheny Bar Association ($1427.55 and $804.30), our personal view *326would be to regard this as a doubtful item (see 388 F.Supp. at 1276) but inasmuch as the Court of Appeals has not displayed an equally skeptical view of the IRS position which regards reimbursement as fatal to the claim of normal club use, we shall count the Smith items as outside income, since we can take judicial notice that members of the bar association pay their own way, and only the newly-naturalized citizens, law school graduates being admitted to the bar, and divers dead-head judges are genuine guests in the IRS view of guest-host relationship. However, the events sponsored by Jack Spinks for L. B. Foster Co. ($877.67) and by Earl M. Keim for Duquesne Light ($755.66) are probably business-related.
According to this allocation, in 1969 the inside business was $23,506.34, or 78.2 percent, and the outside business $6,557.66 or 21.8 percent. Applying this ratio to the previously established banquet receipts, the figures are as follows:
Year Gross receipts Restaurant & Bar Banquet Receipts Inside Outside
1967 477,504 87,975 68,796.45 19,178.55
1968 537,005 145,087 113,458.03 31,628.97
1969 571,482 169,879 132,845.38 37,033.62
1970 620,676 214,700 167,895.40 46,804.60
1971 696,521 231,834 181,294.19 61,439.81
2,903,188 849,475 664,289.45 196,085.55'
Accordingly, we find that the amount of outside business for the years involved is $196,085.55.
2. Percentage of gross receipts
Dividing $196,085.55 by $849,475 we find that the percentage outside business constitutes of the entire gross receipts from the banquet business7 is 23 percent.
3. Amount of “net profits” from outside business
To determine the amount of “net profits” from outside business, according to the Court of Appeals formula, it will be necessary to eliminate from the accountants’ calculations in PX-C-2 through C-6 such amounts as represent allocation of overhead or constant costs so as to ascertain merely the variable or out-of-pocket costs attributable to the “outside business.”8
The categories of expense used by the accountants, in accordance with generally accepted accounting principles, included cost of goods sold, wages and salaries, rent, depreciation, insurance, taxes, employees insurance, supplies and expenses, utilities, maintenance, and miscellaneous.
Of these, cost of goods sold is a proper item, allocated in accordance with the volume of inside and outside business. Wages, employee taxes and insurance, would be proper only if it is proved what employees were hired exclusively for handling outside business. Theoretically, utilities might be *327included also, if it were shown, for example, that certain rooms are heated and lighted only when needed for outside business. However, it is believed that the constant use of facilities available for inside business, as well as outside business, probably makes impracticable such a segregation of expenses.
For simplicity, therefore, we shall calculate merely the cost of goods sold.
This is shown as follows:
Cost of goods sold Banquet Business Outside Costs Outside Revenue “Net Profits”
1967 25.870.71 5,639.81 19,178.55 13,538.74
1968 45,450.07 9,908.12 31,628.97 21,720.85
1969 66,365.68 14,467.72 37,033.62 22,565.90
1970 69,225.31 15,091.12 46,804.60 31,713.48
1971 93.526.72 20,388.82 61,439.81 41,050.99
300,438.49 65,495.59 196,085.55 130,589.96
The amount of “net profits” according to the formula used is therefore $130,589.96 for the five year period, or 15 percent of banquet receipts and 4 percent of gross receipts from restaurant and bar.
4. Purpose for use of club by nonmembers
The purpose of events at which the club facilities are used by outsiders is principally for social events such as weddings or bar mitzvah receptions, and business-oriented gatherings for sales presentations and the like, in connection with the trade, profession, or business of the sponsoring member.
The overwhelming number of business gatherings are for the normal and legitimate business purposes of the sponsoring members, whether to discuss particular developments in the trade, or to promote good will by entertaining past and prospective customers in accordance with standard business custom and practice. Only a few instances come to mind where a member simply served as sponsor for a nonmember desiring to use club facilities for his own purposes. Occasional events such as meetings of the Institute of Architects or the Kaufmann Glee Club or the Duff or Wheeler business schools are out of the mainstream of club activities, which usually center about journalism, advertising, news, public relations, and the particular business activities of the sponsoring members, or purely social gatherings of personal concern to the member holding the affair.
5. Frequency of use by nonmembers
The frequency of events in which nonmembers are entertained is no more than normal in a club of plaintiffs size, 3550 banquet department events having been held in the years 1967-1971 (PX-C-1) of which only 815 were questioned by the IRS. As hereinabove determined, the percentage of outside volume of business amounted to 6.75 percent of total restaurant and bar service.
It is clear that the Press Club is operated in a normal and customary manner, confined to “the service a club usually provides its members and their guests,” and is not a sham or facade for a commercial operation such as a proprietary “bottle club” serving the general public. 388 F.Supp. 1274-75.
Conclusion
In view of the foregoing findings, our determination on the merits is in favor of *328plaintiff. Particularly with respect to the period after the Act of 1969 made unrelated income taxable, so that receipt of such income would not necessarily deprive a club of its exemption, the plaintiff’s exempt status is clear.9 Moreover, the 5 percent criterion favored by the IRS is not necessarily conclusive on the courts. 536 F.2d at 575. The 6.75 percent figure here calculated does not seem excessive. Plaintiff argues that no court decision has upheld revocation where less than 46 percent was shown. 426 F.Supp. at 555.
Retroactivity
The Court has been requested by the parties to pass upon the question of retroactive revocation, whether or not a decision in favor of defendant on the merits is made. On this question of retroactivity our view is that it is not appropriate in the case at bar. Plaintiff has not been guilty of misrepresentation or operating in a manner materially different from its professed intentions.
The club’s letter of June 3, 1964 (GX-4) we continue to believe should be interpreted as referring to the abandonment of the prior practice of permitting use of facilities by completely unrelated groups, rather than member-sponsored “outside” groups. 388 F.Supp. at 1276-77. Hence there was no deliberate and intentional deception which would warrant retroactive revocation. See Lesavoy Foundation v. CIR, 238 F.2d 589, 593 (C.A.3,1956). Moreover, as pointed out at 579 F.2d at 763, the restrictive criteria now applied by the IRS were not developed until later in 1964 and 1971 and were not propounded by the IRS at the 1964 audit of the Press Club. In the event our determination that the plaintiff is entitled to exempt status for the years in question before 1971 as well as for 1971 is erroneous, revocation should be effective only prospectively from February 2, 1972, when the IRS made its determination.
This opinion shall be deemed to constitute the Court’s findings of fact and conclusions of law.
Counsel for the parties are directed to compute and present to the Court a judgment in accordance with this opinion.
. For prior proceedings, see Pittsburgh Press Club v. U. S. 388 F.Supp. 1269 (W.D.Pa.1975); 536 F.2d 572 (C.A. 3, 1976); 426 F.Supp. 553 (W.D.Pa.1977).
. Only business of the “banquet department” need be considered as all parties of ten or more must be handled by that department because the regular dining room and bar does not have sufficient space for large groups. 426 F.Supp. at 554.
. Tr. II, 29. Tr. II 24, gives $145,089.
. Ex. C-6 gives this figure and also $169,878. We arbitrarily selected variants to make the total $849,475.
. We counted as doubtful those where the member answered “no” to the question “Were charges . . . not reimbursed” and “yes” to the question “Were all persons . . . there for your personal, social, or business benefit” but did not disclose who was the employer or person reimbursing the member, or what the relationship of the member to the person reimbursing was. Examples of outside income are where the member answered “no” to both the above questions, and to the question as to business, personal or social purpose of members served by non-member’s use answered “None (Served ás non-member’s sponsor for use of Club facilities).”
. The exhibit is often illegible, and names may be erroneously deciphered.
. This assumes that “gross receipts from nonmembers” means gross receipts from the banquet business. If the entire bar and restaurant business is taken into consideration, the percentage would be lower, i. e. 6.75 percent.
. With respect to “inside” business it would seem proper to utilize total costs, for “banquet business” as well as for other restaurant and bar business. However, for ease of computation and to give the IRS the benefit of the doubt, giving effect to the duty of the taxpayer to keep adequate records substantiating benefits claimed, we shall not attempt to differentiate between the types of banquet business in eliminating fixed costs.
. See 536 F.2d at 575 and 579 F.2d at 762. This would cover 1971.