H. Hollander Co. v. United States

FORD, District Judge.

This is an action for the recovery of penalties and interest alleged to have been wrongfully imposed under the internal revenue laws. Plaintiff is a Massachusetts corporation engaged in Boston in the sale of spirituous liquors, and during the year 1941 was engaged solely in the retail sale of such liquors. Cyril M. Hollander is president and treasurer of the plaintiff.

The Revenue Act of 1941 imposed a floor stocks tax on distilled spirits, § 533, 26 U.S.C.A. § 2800, and on wines, § 534, 26 U.S.C.A. § 3192. In accordance with regulations under the act, plaintiff, on October 1, 1941, had an inventory made of its liquors subject to the tax, and filed a return thereon on December 30, 1941, and on that date paid to the Collector of Internal Revenue the tax of $1417.44 shown to be due by the return. On April 3, 1942, the District Supervisor of the Office of Collector of Taxes wrote plaintiff that the return showed an overpayment of $7.37. On April 14, 1942, plaintiff filed a claim for a refund of this amount. This claim was rejected on August 8, 1944, at which time, as will appear, the United States claimed there had been an underpayment rather than an overpayment.

During September, 1942, two investigators from the Alcohol Tax Unit visited plaintiff’s place of business and stated that the 1941 floor stocks tax return was under investigation. Plaintiff furnished them with such records as they requested, including records of cash receipts and disbursements, the general ledger and invoices for the years 1940-1942 and the December, 1941, inventory. In October, 1942, and January, 1943, Hollander went to the offices of the Alcohol Tax Unit for conference with agents there at the request of the District Supervisor.

On July 7, 1944, after the tax officials had made a determination that the amount of liquors held by plaintiff on October 1, 1941 was greater than that shown on its inventory and return, an assessment of deficiency was made in the amount of $3523.29, of which $1876.38 was the alleged deficiency itself and $1646.91 was an additional 50% of the total tax, assessed under 26 U.S.C.A. § 3612(d) (2) for “a false or fraudulent return or list * * * wilfully made”. First notice of deficiency and demand of the unpaid balance was dated July 15, 1944. A claim for abatement, filed by plaintiff on July 18, 1944 was rejected on July 24, 1944. A second notice and demand dated September 20, 1944 added to the amounts contained in the previous notice a 5% penalty of $176.16, 26 U.S. C.A. § 3655(b), and delinquency interest of $38.12, making the total claim $3757.-57. Another notice and demand, dated February 21, 1945, also designated a *179second notice, included interest to that date of $126.77, making the total amount claimed $3826.22. Plaintiff disputed that any additional amount was due and made no payment until October, 1945, when a distraint warrant was presented by the Collector. Plaintiff then, on October 8,1945, paid the claimed additional floor stocks tax of $1876.38 but not the. other amounts claimed, believing that these claims could then be compromised for a nominal amount.

No further action took place until June 15 and 20, 1950, at which time the Collector distrained from sums on deposit in plaintiff’s bank the sum of $2607.21, made up of the $1646.91 assessed under 26 U.S.C.A. § 3612(d) (2), less a credit of $11.84 on account, the 5% penalty of $176.16 and delinquency interest to that date of, $796.48.

On December 26, 1951, plaintiff filed a claim for refund of the amount of $2607.21, which was rejected by the Bureau of Internal Revenue on September 23, 1952. Plaintiff’s attorney protested this rejection by letter on October 8, 1952, and by letter of October 28, 1952 the Commissioner affirmed the rejection. On March 11, 1953, plaintiff’s attorney was in Washington for conferences with officials of the Bureau of Internal Revenue on the matter, and, thereafter, on April 7, 1953, the Commissioner by letter stood on the rejection. On April 15, 1953, plaintiff commenced its present action for recovery of the $2,607.21.

The sole issue in the case is whether the plaintiff’s inventory of October 1, 1941 and the tax return, based thereon, understated the amounts of taxable liquors held by plaintiff on that date, and, if so, whether such misstatement was a false and fraudulent return wilfully made.

On behalf of plaintiff, Cyril M. Hollander testified that on October 1, 1941 a true and correct inventory of the plaintiff’s liquor stocks had been made under the direction of its manager, James A. Morse, since deceased. The actual physical count was made by two employees, Poster and Ahern, neither of whom is presently employed by plaintiff. Each testified that he had made and entered on work sheets a correct count of the stock in the part of the store assigned to him. The regular inventory sheet was prepared by Hollander, who copied the information correctly from the work sheets. The tax return prepared from these inventory sheets was signed by Morse, who, on December 30,1941, swore that it was a true and correct return.

Hollander further testified that on October 9, 1941 one Vaughan, an inspector from the Alcohol Tax Unit, visited plaintiff’s place of business, examined the inventory, made a spot check on the stock, and then signed a verification of the inventory. Vaughan denied making any check on the stock and said that his verification merely indicated that he had examined and copied into his notebook the totals shown on the inventory sheets. Vaughan originally testified that he had carried out his inspection in accordance with written instructions given to him. Later a copy of these instructions was introduced by plaintiff, directing the agents that, in verifying inventories, they should, without checking each individual item, “be reasonably certain that all taxable liquors are reported on the inventory.” Vaughan, when recalled, testified that on October 9, for the first time, he had been given oral directions by his superiors to omit any actual check on the stocks. This statement was not convincing. On this testimony the court finds that the version of the facts given by Hollander is correct and that Vaughan did make an actual physical check of the stock in plaintiff’s premises as of that date. Prom this, it must be concluded that on October 9 the stock sufficiently corresponded to the October 1 inventory figures to create no suspicion on the part of Vaughan that these figures represented, as the United States now contends, only about half the stock on hand on October 1,

To establish that the actual liquor stocks of plaintiff on October 1, 1941 were greatly in excess of the amounts shown on plaintiff’s inventory of that *180date, the government relies on computations made by its accountants, principally by Raymond Salka who testified as to how they were made. Liquors subject to the floor stocks tax were divided into five categories, in accordance with the different tax rates imposed — whiskey, brandy, wine of less than 14% alcoholic content, wine of over 14% alcoholic content, and champagne. The percentage of liquors in each category in the October 1, 1941 inventory, the December 31, 1941 inventory, and the plaintiff’s purchases during October, November, and December of 1941 (as verified by information furnished by plaintiff’s suppliers) was determined and the average of these percentages taken. These percentages were then applied to the sum of $98,440.-89, representing the total cash receipts of plaintiff for the same three-month period to break this total down into sums representing the cash sales in each category. The difference between the average cost of a case of liquor in each category, as shown by the wholesalers’ records, and the average selling price of a case, as shown by plaintiff’s inventory of December 31, 1941, was taken as the average profit per case. This profit, divided by the cost, gave the markup in percentage. The total cash sales in each category, divided by the cost (100%) plus the percentage of mark-up, gave the cost of goods sold for each category. This figure, divided by the average cost per proof gallon,1 gave the number of proof gallons sold. The number of proof gallons shown on the October 1 inventory, added to the number of proof gallons purchased during the succeeding three months gave the total proof gallons available for sale. From this figure was subtracted the computed number of proof gallons sold. The result should be the number of proof gallons on hand on December 31, if the October 1 inventory was correct. Where the actual December 31 inventory showed a larger figure this was taken to indicate a tax evasion on the theory that the excess indicated an amount of liquor actually on hand October 1, but not shown on the inventory of that date. A similar computation was made using the sum of $12,549.06, which was the total of accounts receivable as shown on plaintiff’s books as of December 31, 1941 and was taken to represent sales on accounts receivable during the preceding three months. These sales, converted into proof gallons, were taken to represent additional evasions, and combined with the amounts from the cash sales computation to arrive at the total evasion charged by the government. These evasions for the various categories, as shown on defendant’s exhibit C, summary of the computations, were divided as follows:

Accounts

Cash receiv-

sales able Total

Whiskey $ 829.13 $785.56 $1614.69

Brandy 191.66 51.27 242.93

Wine 0-14% 6.25 1.47 7.72

Wine 14-21% 27.49 Champagne 153.00 9.80 37.29 32.68 185.68

$1207.53 $880.78 $2088.312

There are several fallacious assumptions which tend to vitiate the accuracy of the results of these computations. The average percentages used to break up total sales into sales within each category were obtained by averaging three dissimilar items — two inventories, and a total of purchases of a three-month period. One of these, the October 1 inventory, did not contain in its prices the new tax which was contained in the other two. There is of course some correlation between sales and inventories. But the *181relationship is not necessarily so close that it can safely be assumed that the averages obtained in these computations will give a fully accurate picture of the sales during the three-month period in question. The computations used a single mark-up for each category based on an average cost per case and an average selling price per case. There is nothing to show that any consideration was being given to weighting these averages to take into account the differences in markup which existed between different brands and qualities within each category and the distribution of sales between brands having a high mark-up and those having a low mark-up. It appears to have been assumed that there was an even distribution of sales throughout the various price ranges in each category, and also that the December 31 mark-up prevailed uniformly throughout the three-month period.

. It is not necessary, however, to try to determine to what extent these assumptions of the calculators may have affected the accuracy of the final result, since there is a more significant error which completely destroys the value of these computations. To understand this, it is necessary to indicate briefly the plaintiff’s method of handling its sales. When liquor was sold for cash, the amount received was rung up on the cash register and no other record made of the sale. When payment was not made at the time of sale, a sales slip was made out, and placed in a file which constituted the company’s accounts receivable ledger. As payment was made on these accounts, the payment was noted on the sales slip and the money received was rung up on the cash register. Thus the cash receipts for any given period represented a total of the cash sales during that period and the payments made during that period on accounts receivable, regardless of when the sale was made which gave rise to the receivable item.

The validity of the government’s computations is based on the assumption that the $98,440.89 figure represents plaintiff’s actual cash sales during the last quarter of 1941 and the figure of $12,-549.06, the sales on accounts receivable for that period. While it may have been, as Salka testified, that he asked for and thought he had received from plaintiff’s accountants the cash sales total for the three-month period, the $98,440.89 actually was the total cash receipts for the period. The figure differs by only thirty cents from the total sales figures of $98,441.19 shown in plaintiff’s ledger for that period. This figure, as has been explained, includes all cash receipts and not just cash sales. Hence, the figure used in the computations as representing cash sales for the period from October 1, 1941 to December 31, 1941 included also amounts received in payment for liquor sold on account before October 1, 1941.

The magnitude of the error here can be seen from the fact that on September 30, 1941 plaintiff’s outstanding accounts receivable totaled $27,284.36, a figure which was unusually high because of heavy purchases by customers just before the new tax rates went into effect on October 1. On December 31, 1941, the total outstanding accounts receivable had been reduced to $12,549.06, which the government assumed represented sales on accounts receivable during the last three months of 1941. This assumption is not necessarily valid, since part of this amount may represent accounts receivable outstanding on October 1 and still unpaid on December 31.

An accurate determination of plaintiff’s sales for the three months in question must take into consideration the accounts receivable outstanding on October 1, 1941. The government’s assumption was that the total sales amounted to $110,989.95 composed of cash sales of $98,440.89 and sales on accounts receivable of $12,549.06 during that period. The first of these component figures does, of course, include all cash sales for the period, and also sales on accounts receivable made during that period, and actually paid for by December 31. The sales on accounts receivable made during the three-month period and not paid by *182December 31 are, of course, included in the second of the component figures. But the $27,284.36 in accounts receivable outstanding October 1 is also included in one or the other of these component figures. That part of this amount paid between October 1 and December 31 is included in the $98,440.89 of cash receipts. The rest which was still unpaid as of December 31 is included in the $12,549.06 representing the total of accounts receivable outstanding as of that date. The whole of the $27,284.36 was thus improperly included in the government’s figures, and the correct total of plaintiff’s sales, whether for cash or on accounts receivable during the three-month period was only $83,705.59. In other words, of the total amount of sales which the government considered the plaintiff to have made, 24.58% was wrongly included, since it represented liquor sold prior to October 1, 1941.

The significance of this can be illustrated by applying this figure to correct the computations for the whiskey category, which shows most of the tax evasion according to the government’s calculations. Combining the government’s computed sales of 6162.33 proof gallons for cash and 785.56 on accounts receivable, we get a total of 6947.89 proof gallons allegedly sold by plaintiff in the three-month period. Subtracting this from the 8105.27 proof gallons available for sales left a total of 1157.38 proof gallons which should have been on hand on December 31. This, compared with the actual year end inventory of 2772.07 proof gallons, gave a computed evasion of 1614.69 proof gallons. Reducing the total computed sales of 6947.89 proof gallons by the 24.58% improperly included in sales for the period, we get a corrected sales figure of 5240.10 proof gallons, and, consequently, a computed December 31 inventory of 2865.17 proof gallons — ■ larger than the actual inventory and hence giving no indication of an understatement on the October 1 inventory. In view of the fact that, as previously pointed out, that the assumptions made in dividing sales among the various categories of liquor and in determining markup on each category can be expected to lead to inaccuracies, this corrected figure is so close to the actual inventory figure of 2772.07 proof gallons as to give strong-confirmation of the correctness of plaintiff's October 1 inventory.

The government’s computations, properly corrected to take into consideration plaintiff’s outstanding accounts receivable on October 1, 1941. offer no adequate support to government’s claim of wilful tax evasion, but. rather tend strongly to corroborate plaintiff’s contentions. On the evidence, it. must be found that plaintiff’s inventory of October 1, 1941, in all probability, was-true and correct and on all the evidence-it is concluded that the plaintiff did not* either wilfully or otherwise, make any fraudulent or false return or list; that, the taxes sought to be recovered were-wrongfully assessed and collected; and' that the plaintiff is entitled to recover the sum of $2607.21 with interest at 6%-from June 20, 1950.

Judgment for plaintiff in accordance; herewith.

. Proof gallon was the taxable unit for whiskey and brandy. In computations for the other categories the appropriate taxable unit for each was used instead of the proof gallon-wine gallon in the case of wine, and % pint in the case of champagne.

. Exhibit O shows a total evasion of $2,088.31. The evasion claimed and collected by the government was $1876.38. No explanation was offered by the government as to this discrepancy or as to how the $1876.38 figure was arrived at.