GOLDBERG, Circuit Judge:
We have for review a two-pronged decision of the Tax Court. Bolen Webb and Cornelia Webb, T. C. Memo 1966-81 (April 20, 1966). One prong relates to deficiencies in reporting income tax liability, and the other relates to their sometimes consort, the 50% fraud penalty. The taxpayers, Bolen Webb and his wife Cornelia Webb (hereinafter Webb), seek our disavowal of the Tax Court’s findings and conclusions on both prongs. We see no reason for such disavowal as to either the deficiency or the fraud and affirm.
The years during which Webb’s income is in question are 1958, 1959, and 1960. The aggregate dollar amounts in question are $30,076.34 for the deficiency and $15,038.17 for the concomitant fraud penalty. 26 U.S.C. § 6653(b).
Webb has been in the retail liquor business in Dallas, Texas, since 1946. During the years involved in this controversy, 1958 to 1960, he operated three retail liquor stores; and in July, 1960, he began operating a fourth store.
The Webbs’ first encounter with the Commissioner occurred in 1958, when he challenged their joint returns for the years 1956 and 1957. These returns had reported income from Webb’s three liquor stores as follows:
1956 1957
Sales $76,713.00 $65,701.90
Cost of Sales 57.450.00 46,949.70
Gross Profit 19.263.00 18,752.20
Expenses 19,746.97 18,212.00
($ 483.97) $ 525.20
The returns were investigated by Internal Revenue Agent Fred A. Bush between August of 1958 and June of 1960. Finding Webb’s books and records inadequate, Agent Bush reconstructed his income, using a 20% mark-up on the known cost of sales. As Bush later testified, the 20% mark-up was a compromise adjustment agreed upon by Bush, Webb, and J. L. Lobdell who operated a tax and bookkeeping service and who had prepared Webb’s 1956 and 1957 returns. Bush thus reconstructed Webb’s Net Profit and Unreported Net Income for the years 1956 and 1957 as follows:
1956 1957
Sales $128,861.32 $141,538.02
Cost of Sales 107,384.44 117,948.35
Gross Profit 21,476.88 23,589.67
Expenses 10,608.28 13,088.89
Net Profit 10,868.60 10,500.78
Net Profit Reported ( 483.97) 525.20
Unreported Net Income $11,352.57 $ 9,975.58
Based on these figures, the Webbs were assessed the following additional taxes and a 50% fraud penalty:
1956 1957
Tax Deficiency $1,760.46 $1,654.60
Fraud Penalty under 26 U.S.C. § 6653(b) 880.23 827.30
Total $2,640.69 $2,481.90
Webb signed a Form 870 agreement (Waiver of Restrictions on Assessment and Collection of Deficiency in Tax and Acceptance of Overassessment) regarding the assessment for the years 1956 and 1957.
The prophetic sting of the $5,122.59 tax assessment was reenforced by several warnings for Webb to keep better records. In June of 1959 Agent Bush advised Webb that the Internal Revenue laws require all taxpayers to keep sufficient records and that Webb should keep in permanent form a daily record of the money which he took in and of all purchase invoices and receipts. A month later the District Director of Internal Revenue in Dallas, Texas, sent Webb a letter advising him of his duty to keep accurate and complete business records and reprimanding Webb for his past failure to do so. The letter reads as follows:
“[Y]our current records were found to be inadequate in the following respects: No books and records of any sort were maintained. Receipts were not deposited or recorded. Inventories were based on loose estimates rather than actual physical count and valuation. Rental income and expenses were not recorded.
* * * * * •>:•
Each taxpayer is required by law to make a return of his true income. He must, therefore, maintain such accounting records as will enable him to do so. * * *
******
This notice to you of the requirements of the law and regulations with respect to maintaining such permanent books of account and records * * * as will enable you to make a return of your true income and also will enable examining officers of the Internal Revenue Service to determine the correct amount of income subject to tax.”
Moreover, on several occasions in 1958 and 1959 Lobdell talked to Webb about the necessity of maintaining adequate records. He recommended that Webb keep a detailed record of receipts for all expenditures and purchases and a daily record of all sales from each store.
Despite the above presagements Webb refused to alter his bookkeeping in any significant manner. In the latter part of 1958 he did begin keeping part of his invoices in a whiskey box. However, some of the invoices were lost or misplaced, so that he was unable to furnish the revenue agent who conducted a subsequent examination with a complete record of his purchases during 1958, 1959, and 1960. During 1960 Webb kept a “cash receipts book” containing weekly entries of his sales for that year. Not only was the book incomplete, containing entries for only 48 weeks of the year 1960, but also it was totally disregarded in determining Webb’s taxable income for 1960. The
gross
sales which Webb reported for the taxable year 1960 were approximately $18,000
less
than the gross sales recorded in the admittedly incomplete “cash receipts book.”
Webb did not ring up all cash sales on the cash registers of his business, and he did not deposit all of his receipts in bank accounts. Although he maintained accounts in one bank during the three-year period and in a second bank for six months during the period, he was unable to locate all of his canceled checks and other banking records for 1958, 1959, and 1960.
The Webbs’ joint income tax returns for 1958, 1959, and 1960 reported income from their retail liquor business as follows:
1958 1959 1960
Sales $101,218.78 $101,121.62 $110,381.12
Cost of Sales
91,345.05 91,000,58 100,150.70
Gross Profits 9,873.68 10,121.04 10,230.42
Expenses 10,001.13 10,008.60 9,622.36
Net Profit $( 127.45) $ 112.44 $ 608.06
These returns reported the following inventories and inventory purchases, with the resulting cost of sales:
1958 1959 1960
Opening Inventory $10,000.00 $10,000.00 $ 10,000.00
Inventory Purchases 88.419.30 87.551.00 97,670.00
Total 98.419.30 97.551.00 107,670.70
Ending Inventory 10,000.00 10,000.00 12,000.00
Costs of Sales $88,419.30 $87,551.00 $ 95,670.70
In September of 1961 Revenue Agent Larry Kelley was assigned to investigate the Webbs’ 1958, 1959, and 1960 returns. He found Webb’s books and records to be totally inadequate. Likewise, after examining Webb’s canceled checks and his primary bank’s microfilm of Webb’s canceled checks, bank statements, and deposit slips, Agent Kelley was no more enlightened. He thus estimated Webb’s liquor income by the percentage mark-up method, as had been done regarding the 1956-57 settlement.
Kelley examined the records of all of Webb’s wholesalers and found that Webb had reported about one-half of the actual purchases on his tax returns. His actual inventory purchases had been as follows:
1958 1959 1960
Liquor and Wine $151,773.13 $173,006.08 $179,055.34
Beer 12,949.16 16,417.88 23,641.60
Total $164,722.29 $189,423.96 $202,696.94
There was also a discrepancy between the opening inventory reported on the 1958 return and the closing inventory agreed to in the 1956-57 settlement; therefore, Kelley recomputed Webb’s cost of sales using both the new purchase figures and the opening inventory determined by the 1956-57 settlement. His computations resulted in the following statement of cost of sales:
1958 1959 1960
Opening Inventory $ 13,000.00 $ 10,000.00 $ 10,000.00
Inventory Purchases 164,722.29 189,423.90 202,696.94
Total $177,722.29 $199,423.96 $212,696.94
Ending Inventory 10,000.00 10,000.00 12,000.00
Cost of Sales $167,722.29 $189,423.96 $200,696.94
Using a 25% mark-up from cost of sales (the propriety of which will be discussed infra) Kelley determined Webb’s net profit and unreported income as follows:
1958 1959 1960
Sales (125% of Cost of Sales) $209,652.86 $236,779.95 $250,871.15
Cost of Sales 167,722.29 189,423.96 200,696.92
Gross Profit 41,930.57 47,355.99 50,174.23
Expenses
12,376.88 13,673.87 16,965.56
Net Profit 29,553.69 33,682.12 33,208.67
Net Profit Reported ( 127.45) 112.44 608.06
Unreported Net Income $ 29,681.14 $ 33,569.68 $ 32,600.61
The Commissioner assessed Webb for both the tax deficiency and a 50% fraud penalty:
1958 1959 1960
Tax Deficiency 9,413.74 $10,101.14 $10,561.46
Fraud Penalty Under 26 U.S.C. § 6653(b) 4,706.87 5,050.57 5,280.73
$14,120.61 $15,151.71 $15,842.19
The total monetary demand on Webb was $45,114.51.
Before the Tax Court, as'before this Court, Webb attacked the assessment in three areas: (1) the use of the percentage mark-up without regard for net worth considerations, (2) the use of a mark-up percentage of 25% for determining gross sales, aud (3) the ascription of fraud to Webb in the filing of his tax returns. Our analysis will focus on these three areas.
(1) Use of the Percentage Mark-up Method sans Net Worth Considerations
It is conceded that Webb had no adequate books and records to reflect his income during the contested years. Webb contends, however, that a net worth analysis of his financial status in 1958, 1959, and 1960 shows that he could not possibly have earned $96,444.48, which the Commissioner, using the percentage mark-up method determined to be his total liquor business income. In Webb’s brief he states in part, “Ignoring the undisputed testimony as to Petitioner’s net worth was clear error. * * In no other percentage mark-up case have considerations of net worth been totally disregarded.”
Our analysis starts with the basic principle that Webb, along with every taxpayer, must maintain accounting records which enable him to file a correct tax return. 26 U.S.C. § 6001; Section 1.446-1, Income Tax Regs.
Because Webb’s records did not clearly reflect his income, the Commissioner was authorized to use such methods as in
his opinion clearly reflected that income. 26 U.S.C. § 446(b).
The determination of taxable income by the Commissioner is presumptively correct. Anson v. C.I.R., 10 Cir. 1964, 328 F.2d 703, 706; Breland v. United States, 5 Cir. 1963, 323 F.2d 492, 496; Mendelson v. C.I.R., 7 Cir. 1962, 305 F.2d 519, 522, cert. den., 371 U.S. 877, 83 S.Ct. 149, 9 L.Ed.2d 114; Klassie v. United States, 8 Cir. 1961, 289 F.2d 96, 100. See also Welch v. Helvering, 1933, 290 U.S. 111, 54 S.Ct. 8, 78 L.Ed. 212; Broadhead’s Estate v. C. I. R., 5 Cir., 391 F.2d 841, March 5, 1968. The Tax Court’s approval of that determination will not be set aside by this Court unless clearly erroneous. 26 U.S.C. § 7482(a); Fed. R.Civ.P. 52(a)
;
Browne v. C.I.R., 4 Cir. 1966, 367 F.2d 386, 387; Anson v. C.I.R., supra, 328 F.2d at 706-707; Agnellino v. C.I.R., 3 Cir. 1962, 302 F.2d 797, 799. See also C.I.R. v. Duberstein, 1960, 363 U.S. 278, 289-292, 80 S.Ct. 1190, 4 L.Ed.2d 1218, 1227-1229; Estate of Sam E. Broadhead v. C.I.R., supra. Especially when the taxpayer has defaulted in his task of supplying adequate records, he is not in a position to be hypercritical of the Commissioner’s labor. As we said in Bernstein v. C.I.R., 5 Cir. 1959, 267 F.2d 879, 881:
“It is immaterial whether the Commissioner proceeded upon a wrong theory in determining the deficiency, and the taxpayer has the burden of showing that the assessment is wrong on any proper theory.”
Webb’s broad assertion that net worth determinations are always considered is not supported by the cases. In Kurnick v. C. I. R., 6 Cir. 1956, 232 F.2d 678, for example, the Sixth Circuit affirmed a deficiency and fraud assessment under the percentage mark-up method without mentioning net worth considerations. Eleven years later the same Circuit upheld a deficiency and fraud assessment despite the taxpayer’s contention that his net worth computations had successfully countered the Commissioner’s percentage mark-up estimate. The Tax Court had dismissed the contention by noting that “the petitioners did not introduce any evidence to show the correctness of the items contained therein.” Carmine Bollella, 24 T.C.M. 858 (1965), aff’d., Bollella v. C.I.R., 6 Cir. 1967, 374 F.2d 96. Although in the case at bar Webb did introduce “testimonial evidence” as to his net worth, we question whether his self-serving statements satisfy the quest of the
Bollella
courts. We refer to Miller v. C. I. R., 5 Cir. 1956, 237 F.2d 830, where the Commissioner’s reconstruction of income under the “bank deposit method” was challenged by a net worth attack. As in this case, the taxpayer’s integrity had been placed in doubt, and we concluded as follows:
“The taxpayer complains that the agent should have followed the increase in net worth method of computation.
* * * There can be no requirement that the Commissioner use for his calculation of income a method which depends absolutely for its accuracy on the taxpayer’s voluntary disclosure of his assets at the end of the period in a case where the court rejected his contention that he had a large amount of cash at the beginning of the period and when he overlooked a bank account of $5,000 at the end of the period.” 237 F.2d at 838.
For Webb the percentage mark-up method was the most reasonable means of computing income because his cost of sales in each taxable year was known with certainty. In comparison, net worth or bank deposit computations would lack any known figure because Webb transacted a substantial amount of his liquor business in cash, and he readily commingled cash flowing to and from personal and business expenses. The supposed disparity between the percentage markup and net worth determinations is based almost exclusively on Webb’s own testimony concerning his property, expenses, and money hoards. Moreover, the Commissioner did disclose possible unreported income drains in the elimination of certain debts and the purchase in 1959 of two new air-conditioned Cadillacs.
We recognize that the absence of adequate tax records does not give the Commissioner carte blanche for imposing Draconian absolutes. Gasper v. C. I. R., 6 Cir. 1955, 225 F.2d 284. See also Helvering v. Taylor, 1935, 293 U.S. 507, 55 S.Ct. 287, 79 L.Ed. 623. But such absence does weaken any critique of the Commissioner’s methodology. Mendelson v. C. I. R., supra, 305 F.2d at 523.
Arithmetic precision was originally and exclusively in Webb’s hands, and he had a statutory duty to provide it. He did not have to add or subtract; rather, he had simply to keep papers and data for others to mathematicize. Having defaulted in his duty, he cannot frustrate the Commissioner’s reasonable attempts by compelling investigation and recomputation under every means of income determination. Nor should he be overly chagrined at the Tax Court’s reluctance to credit every word of his negative wails. See Breland v. United States, supra, 323 F.2d at 496 (at [3-5]); Mendelson v. C. I. R., supra, 305 F.2d at 521 (at [1]).
(2) Mark-up Percentage of 25% for Determining Gross Sales
Predictably, Webb claims that the Commissioner’s mark-up percentage of 25% is arbitrary and unreasonably high. His major argument is that shelf prices, upon which the mark-up percentage was based, are not reliable representations of selling prices because his stores regularly granted liberal discounts. He adds that even selling prices do not accurately report income because of the magnitude of business inefficiency and theft in his stores. The Tax Court considered all of Webb’s contentions and found that a mark-up of 25% was fair and reasonable for Webb’s business during the years 1958 through 1960. After a brief recitation of the relevant facts, we will discuss the eight specific areas in which Webb challenges the 25% mark-up figure. In doing so, however, we emphasize that the same presumptions of verity are accorded to Commissioner and Tax Court determinations here as in Section (1) of this opinion, supra.
Revenue Agent Kelley began his investigation of Webb’s 1958, 1959, and 1960 income tax returns in September of 1961. As has been mentioned previously, Kelley computed the cost of sales by obtaining accurate records from Webb’s suppliers. To arrive at a markup percentage, he first averaged the 1959 and 1960 shelf prices of what, according to Webb, were the eighteen best selling items.
Based on these shelf prices, the
average mark-up on all of Webb’s most popular brands in all sizes was computed to be 34.25% over cost for 1959 and 30.50% for 1960. The average mark-up on these brands in the most frequently-sold half-pint sizes was 36.68% and 32.-64% for 1959 and 1960, respectively. Kelley reduced the average mark-up to 25% to account for “any possible discounts or losses for breakage, theft, minor changes of price during the year.” Using this mark-up of 25%, Kelley reconstructed Webb’s sales for each of the three years and deducted therefrom the costs of sales and business expenses.
The eight specific challenges will be considered separately.
Discount of odd pennies.
J. T. Light-ner, one of Webb’s store managers, testified that he would “knock off the pennies, if the shelf price of the bottle was any price not divisible by five. Although Lightner indicated that this was Webb’s policy for all stores,
he never explicitly stated that, and no other witness — including Webb himself — even mentioned that pennies were discounted at any other store. Moreover,
no
shelf prices in the computations for 1959 included odd-penny amounts. Therefore, we find the “odd-penny discount” computations found in Webb’s brief (which still obtained a mark-up percentage above 25%) to be totally unconvincing.
Quantity discounts.
Webb asserts in his brief: “Every person who ever sold liquor in Petitioner’s store testified that he gave large discounts on quantity sales. * * * ” We need not attack the verity of that statement; we will instead analyze its persuasive power. Both Bolen and Cornelia Webb testified that they granted five to twenty-cent discounts on quantity sales (more than five bottles). Jewel Jackson, a store manager, eorrobo-rated the Webbs and added that he often sold a twelve-bottle case of Thunderbird wine for $7.50 (by 1960 wholesale prices, only a 7.59% mark-up). J. T. Lightner testified that he occasionally made quantity sales at a discount, but only after calling Webb for his approval. J. D. Cooper, who managed a store which was open only six months during the three-year period involved, testified that he “sold a lot of whiskey to bootleggers” without telling Webb. However, he also testified that some of his shelf prices were higher than those used by Kelley in determining Webb’s mark-up. None of the above stated what percentage of their business was in quantity discount sales.
Especially as to the testimony of Light-ner and Cooper, the Commissioner’s allowance for discounts seems generous enough. As to the testimony of the Webbs and Jackson, even if substantial discounting can be inferred, the Tax Court was justified in “discounting” the testimony due to an interest in the proceedings. See Archer v. C. I. R., 5 Cir. 1955, 227 F.2d 270, 273. Jackson, for instance, was still employed by the Webbs at the time of the Tax Court hearing, and his testimony was impeached by Revenue Agent Kelley’s recitation of a prior conversation with Jackson.
General discounting procedures.
It is true that all witnesses testified to the practice of discounting from five to twenty-five cents on certain sales to meet the competition of other stores. Webb’s testimony on this regard is typical:
“Sometime — well, people would come in often and say I am short and would like a nickel having and sometimes fifteen cents having the price and I would say let them have it.”
The same problems of frequency and credibility plagued Webb in this chal
lenge as they did concerning the quantity sales. Moreover, Webb himself testified twice that in more than half his sales no discounts were given. In its appellate brief the government has computed the mark-up percentages (1) if all sales were discounted ten cents from shelf prices and (2) if half of the sales were discounted ten cents from shelf prices. The following computations show that the
25%
general mark-up was a reasonable one:
1959 1960
All Sales Half-Pints All Sales Half-Pints
At Shelf Prices 34.28% 36.68% 30.50% 32.64%
At 10 cents Discount per Bottle 28.50% 27.49% 23.92% 23.70%
At 10 cents Discount in Half the Sales 31.39% 32.08% 27.21% 28.17%
Change in shelf price.
Webb contends that Revenue Agent Kelley failed to take into account any shelf price changes from 1960, the last tax year involved here, to 1961, the year in which Kelley computed mark-up from the current shelf prices. As mentioned in footnote 8 supra and as found by the Tax Court, the 1960 and 1961 shelf prices were substantially the same. The Tax Court granted that “[t]here had been a minor shelf price change in 1960.” However, Webb did not show how substantial this “minor change” was.
Unweighted mark-wp.
Webb criticizes Revenue Agent Kelley for making “no effort to arrive at a weighted average” of the eighteen items used to compute the mark-up percentage. The Tax Court found that the failure to weigh was due to the difficulty of assigning weights to the $500,000 worth of liquors, beers, and wines and the thousands of bottles of different brands which Webb sold. The Tax Court also noted that the average profit on half-pint bottles, which constituted most of Webb’s sales, was larger than the average profit of all the bottles. We add that Webb on appeal has shown neither his cooperation in seeking a weighted average nor injury to him in Kelley’s failure to compute a weighted average.
Inefficiency and theft.
Webb asserts that the 25% mark-up does not provide for substantial losses caused by inefficient operations, employee thefts, and credit losses. We note that Webb claimed on his returns, and was allowed, the following deductions from gross profits for bad cheek losses: $902.46 for 1958, $943.80 for 1959, and $522.96 for 1960. These allowed losses, therefore, could have no effect on the Commissioner’s determination of gross profit or net income. As to other, purely testimonial, evidence of reduced income due to inefficient operations and credit losses, we find no clear error in the Tax Court’s failure to be convinced.
Webb claimed employee theft in the cases of Lightner and Cooper. Having no direct evidence as to Lightner, he attempted to reconstruct Lightner’s income and expenditures to show that Lightner was taking from the till. Webb’s enthusiasm and diligence in this income determination would have been appreciated in the determination of his own income. However, even here there are substantial gaps, and Webb’s willingness to keep Lightner in his employ through the end of 1962 weakens his argument considerably. As to Cooper, Webb presented more direct evidence, the fact that police arrested Cooper for coming out of his store with whiskey “before day.” But this incident occurred in 1964. There is no evidence that Cooper stole substantial amounts during the first six months
of his employment, June through December of 1960.
Under Section 165(e) of the Internal Revenue Code
a taxpayer is entitled to a theft ioss only in the year of discovery and only if he can prove that a theft in fact occurred and will not be reimbursed. Bodzy v. C.I.R., 5 Cir. 1963, 321 F.2d 331, 335; Naegle v. C.I.R., 9 Cir. 1967, 378 F.2d 397. While we do not hold that this section prevents Webb from offsetting the mark-up percentage with specific or specifically recurring losses, it does indicate a reluctance to grant liberal income deductions several years after the fact. See also Marcello v. C.I.R., 5 Cir. 1967, 380 F.2d 494, 496-497; Patterson v. Pizitz, Inc., 5 Cir. 1965, 353 F.2d 267, 269-270, cert. den., 1966, 383 U.S. 910, 86 S.Ct. 895, 15 L.Ed.2d 666. Cf. Anderson v. C.I.R., 5 Cir. 1957, 250 F.2d 242, cert. den., 1958, 356 U.S. 950, 78 S.Ct. 915, 2 L.Ed.2d 844.
Comparison with 1956-57 deficiency.
Webb urges that the Tax Court ignored the conduct of his business in prior years. For the two prior years the parties agreed to a settlement based on a 20% mark-up on costs of sales of $107,384.-44 and $117,948.35. For the years involved in this dispute, Webb’s business expanded considerably and his costs of sales were $167,722.29 for 1958, $189,-423.96 for 1959, and $200,696.94 for 1960. This increased level of business activity with relatively constant business expenses accounts for much of the increase in net income for the year 1958. Moreover, the 20% mark-up for 1956 and 1957 was a figure agreed to in a discussion among Revenue Agent Bush, Webb, and bookkeeper Lobdell. Bush testified that he had determined that 25% was an appropriate mark-up but that he had reduced it to 20% in settlement. He testified that he thought a 20% mark-up was reasonable for the years 1956 and 1957, but he also testified that he made no examination of 1959 costs, did not compute a mark-up for 1959, and did not think that 20% was a reasonable mark-up for 1959. We find no evidence concerning Webb’s business in 1956 and 1957 which seriously disparages Agent Kelley’s determinations for 1958 through 1960.
Comparison with other businesses.
Bill Barnes, whose bookkeeping services were solicited by Webb in 1962, testified that a reasonable mark-up for liquor stores in the area was less than twelve per cent. The testimony mentioned no specific period of time, and, as the Tax Court found, Webb presented no evidence of prices charged in other stores or of a similarity of pricing and discount practices among liquor dealers. Webb’s own shelf prices indicated a much higher mark-up, and he testified that he sold at shelf prices more than half the time. The Tax Court was not clearly erroneous in disregarding Barnes’ testimony.
In conclusion we acknowledge Webb’s plea to view the totality of his business circumstances. In his appellate brief he concludes:
“As to all of the foregoing, the Tax Court holds that all of these items were properly allowed for in the six percent reduction [down to 25 %]— swept under the rug, if you please. It is clear that the 6% reduction does not have the elasticity the Tax Court accords it; it can be stretched only so far, and not in so many different ways.”
This criticism may have surface plausibility, but the fault for failure lies in Webb himself. He has contented himself by shafting darts of criticism, but each dart lacks substantive power. There are X’s in the Commissioner’s
equation, but the formula is reliably correct, and the X’s were not filled in by Webb. The Commissioner’s computations were supported by reason and experience. Webb underestimates his burden if he thinks that, after defaulting in his original duty, he need merely testify to the Commissioner’s lack of precision.
In Cefalu v. C.I.R., 5 Cir. 1960, 276 F.2d 122, 126, this Court stated:
“If the taxpayer shows that part of the Commissioner’s'determination was wrong, it does not destroy the presumption of correctness that attaches to his findings. Anderson v. Commissioner, 5 Cir. 1957, 250 F.2d 242, cer-tiorari denied 356 U.S. 950, 78 S.Ct. 915, 2 L.Ed.2d 844.”
In Miller v. C.I.R., 5 Cir. 1956, 237 F.2d 830, 838-839, this Court granted that the taxpayer had pointed out specific areas where the Commissioner’s recom-putation method had failed to reflect his true income. Nevertheless, we concluded:
“In no case did the actual invoices proven by the taxpayer even approximate the figure allowed him by the Tax Court for any particular year. The court did the best it could under the circumstances, and we cannot arrive at any different figures. On this phase of the decision, therefore, the Tax Court’s computation must stand.” 237 F.2d at 839.
See also Patterson v. Pizitz, Inc., supra; Anson v. C.I.R., 10 Cir. 1964, 328 F.2d 703, 707; Breland v. United States, 5 Cir. 1963, 323 F.2d 492, 496; Mendelson v. C.I.R., 7 Cir. 1962, 305 F.2d 519, cert. den., 371 U.S. 877, 83 S.Ct. 149, 9 L.Ed.2d 114; Archer v. C.I.R., 5 Cir. 1955, 227 F.2d 270, 273-74.
The Commissioner’s 25% base, while perhaps not overly generous, was neither arbitrary nor capricious. Discounts, losses, theft — all the elements claimed by Webb — were taken into consideration. Webb has failed to supply us with the factual primates of further reduction, even in retrospective analysis.
(3) Fraud
The appellate court’s role in reviewing cases of civil tax fraud entails unique, though not exclusive, dualistic frustrations. To affirm for the Commissioner, we must agree that he has proved fraud by “clear and convincing evidence.” Moreover, what he has proved must be fraud, not mere negligence or even gross negligence. Judge Brown has analyzed the tests in Carter v. Campbell, 5 Cir. 1959, 264 F.2d 930, 935-936:
“There is first the matter of what is and what is not fraud. We have recently given our approval, Olinger v. Commissioner of Internal Revenue, 5 Cir. 1956, 234 F.2d 823, 824, to the strong language of Davis v. Commissioner of Internal Revenue, 10 Cir. 1950, 184 F.2d 86, 87, 22 A.L.R.2d 967. ‘Fraud implies bad faith, intentional wrongdoing and a sinister motive. It is never imputed or presumed and the court should not sustain findings of fraud upon circumstances which at most create only suspicion.’ Equally emphatic is that stated for us by Judge Sibley in Mitchell v. Commissioner, 5 Cir. 1941, 118 F.2d 308, 310. ‘Negligence, whether slight or great, is not equivalent to the fraud with intent to evade tax named in the statute. The fraud meant is actual, intentional wrongdoing, and the intent required is the specific purpose to evade a tax believed to be owing. Mere negligence does not establish either.’ These principles have been restated frequently and recently with no recession either from emphatic language or like application. Goldberg v. Commissioner of Internal Revenue, 5 Cir. 1956, 239 F.2d 316, 321; Fairchild v. United States, 5 Cir. 1957, 240 F.2d 944, 947; Eagle v. Commissioner of Internal Revenue, 5 Cir. 1957, 242 F.2d 635, 638; Jones v. Commissioner of Internal Revenue, 5 Cir. 1958, 259 F.2d 300.
See also 10 Mertens, Federal Income Taxation § 55.10.
“Added to this is the burden placed upon the Commissioner. The Code places it expressly upon the Commissioner. 10 Mertens, Federal Income Taxation § 55.18. ‘The burden is upon the Commissioner to prove affirmatively by clear and convincing evidence actual and intentional wrongdoing on the part of the petitioner with a specific intent to evade the tax.’ Eagle v. Commissioner of Internal Revenue, 5 Cir. 1957 242 F.2d 635, 637. 10 Mer-tens, supra, at § 55.16.”
On the other hand, to reverse we must find that the Tax Court’s determination of fraud was “clearly erroneous” as to
all
items which comprise the deficiency. See footnotes 1 and 7, supra. We quote Judge Jones in Toledano v. C.I.R., 5 Cir. 1966, 362 F.2d 243, 247:
“The statute reads that ‘if any part of any deficiency is due to fraud with intent to evade tax’ the fraud penalty may be imposed and the bar of limitations is avoided. 26 U.S.C.A. (I.R.C. 1939) §§ 293(b), 276(a), now 26 U.S. C.A. §§ 6653(b), 6501(a). Thus, the Commissioner need only prove that some portion of the deficiency is due to fraud. Ginsberg v. Commissioner of Internal Revenue, 5th Cir. 1959, 271 F.2d 511. A finding of fraud by the Tax Court is generally treated as a fact finding which is not to be disturbed unless clearly erroneous. Helvering v. Keho, 309 U.S. 277, 60 S.Ct. 549, 84 L.Ed. 751 (1940); Steiner v. Commissioner of Internal Revenue, 7th Cir. 1965, 350 F.2d 217; Schroeder v. Commissioner of Internal Revenue, 8th Cir. 1961, 291 F.2d 649, cert. den., 368 U.S. 985, 82 S.Ct. 598, 7 L.Ed.2d 523 (1962); Kenney v. Commissioner of Internal Revenue, 5th Cir. 1940, 111 F.2d 374.”
A summary of the above standards demonstrates their tutorial limitations: We must determine whether it is
clearly
erroneous that the taxpayer’s intent to defraud the government was proven, as to any part of the deficiency, by
clear and convincing
evidence. Our path in fraud determinations is even more obstacle-pocked because we have no cinematography of the mind nor do we have books approaching impeccable accuracy. Nevertheless, courts have attempted to avoid deciding each case viseerally and have established certain factual checklists.
The factor most often discussed is the deficiency itself. Although fraud
cannot be inferred from the mere understatement of income, consistent and substantial underreporting is evidence of fraud. Holland v. United States, 1954, 348 U.S. 121, 139, 75 S.Ct. 127, 99 L.Ed. 150, 166; Browne v. C.I.R., 4 Cir. 1966, 367 F.2d 386, 387; Bahoric v. C.I.R., 9 Cir. 1966, 363 F.2d 151, 154; Klassie v. United States, 8 Cir. 1961, 289 F.2d 96, 101-102; Cefalu v. C.I.R., 5 Cir. 1960, 276 F.2d 122, 129. This factor is relevant to our analysis because Webb’s deficiencies in reporting were both consistent and substantial. He reported less than 49% of his sales in 1958, less than 43% in 1959, and less than 45% in 1960. He reported a total of $593.05 taxable business income for the three years while the Commissioner computed his taxable income to be $96,444.48. These are not miniscule mistakes.
Webb counters these figures by reminding us that they are based on reconstructed income determinations. He points out that we must not bootstrap him into a fraud penalty by means of presumed deficiencies. See Goldberg v. C.I.R., 5 Cir. 1956, 239 F.2d 316, 320. This is not a case where the deficiencies are small or where they comprise the total evidence against. Webb. The evidence of Webb’s fraudulently understating
some part
of his income is not only clear and convincing; it is well nigh compelling. In Merritt v. C.I.R., 5 Cir. 1962, 301 F.2d 484, 487, Judge Jones, who also authored the
Goldberg
case supra, brings Webb within the warrant of the Tax Court’s discretion to find fraud:
“The mere understatement of. income, standing alone, is not enough to carry the burden cast upon the Commissioner in seeking to recover fraud penalties. But each case is to be considered in the light of its own facts. Consistent and substantial understatement of income is by itself strong evidence of fraud. This proof, coupled with the showing that the records were both incomplete and inaccurate, and that the petitioner did not supply the bookkeeper with all of the data necessary for maintaining complete and accurate records, is enough to warrant the Tax Court in finding fraud. Reaves v. Commissioner of Internal Revenue, 5th Cir. 1961, 295 F.2d 336; Cefalu v. Commissioner of Internal Revenue, supra; Bryan v. Commissioner of Internal Revenue, supra.”
Once we get over Webb’s abhorrence of the Commissioner’s methodology in determining “how much” was underpaid, it remains undisputed that Webb owed taxes. We can add with certainty that Webb knew his records were inadequate. He knew that the figures he gave to his bookkeepers did not accurately reflect his income because he testified that he did not always record the day’s sales. Moreover, his incomplete “cash receipts book” containing forty-eight weekly entries of his sales during 1960 showed gross sales during 1960 of approximately $18,000 in excess of the gross sales figure on his 1960 tax return.
Nor can we accept Webb’s second defense that any deficiencies are attributable solely to ignorance and perhaps negligence. Iteration and reiteration have convinced us that Webb was short on schooling. He is pictured as untutored and unlearned, unconscious of any guilt, a man with an unburdened conscience. Yet, though we consider intelligence on our scales of culpability, the lack of schooling does not automatically register a zero. In Carmine Bollella, 24 T.C.M. 858 (1965), aff’d, Bollella v. C.I.R., 5 Cir. 1967, 374 F.2d 96, an immigrant with only a first grade education and with no knowledge of bookkeeping was assessed a penalty for civil tax fraud based on a deficiency determined by the percentage mark-up method. (The Tax Court’s refusal to consider Bollella’s net worth computations is mentioned supra.) The penalty was assessed despite Bollella’s lack of formal education and despite the fact that Bollella’s son managed the business, signed checks, received and deposited the business receipts, and was responsible for the preparation of the income tax returns. See also Bahoric v.
C.I.R., 9 Cir. 1966, 363 F.2d 151, 153 (Yugoslav immigrant with only a sixth grade education); Estate of Kalman Bernstein, T. C. Memo. 1956-260, CCH Dec. 22, 029(M), 1956, aff’d, Bernstein v. C.I.R., 5 Cir. 1959, 267 F.2d 879 (immigrant with “equivalent of third or fourth grade education”). Though Webb did not complete the fifth grade, he is a person of average intelligence and average business understanding. In 1958 he was no economic infant or adolescent ; he was a successful businessman doing more than $200,000.00 volume annually. It traduces credibility that his furnishing less than half of his sales to his bookkeeper was due to inadvertence or negligence.
In crossing the thin line between negligence and fraud, we are reenforced by the recollection of Webb’s 1956-57 experience. That which may have been initially negligent can become, after warning, willful and intentional by its continuity. Cf. Holland v. United States, supra, and Hamman v. United States, 9 Cir. 1965, 340 F.2d 145, 149, cert. den., 380 U.S. 977, 85 S.Ct. 1339, 14 L.Ed.2d 271, involving criminal tax fraud. We need not repeat the facts surrounding Webb’s 1956-57 tax assessment and the warnings which followed. Such encounters with the tax authorities would teach one possessing elementary common sense that guessing — and guessing far below the mark — is not permitted. Though the accusing finger had pointed, Webb continued his under-reporting without interval. Ignorance may be a sometime haven, but it is not an all-time refuge from scienter.
The Tax Court found fraud after carefully weighing Webb’s solemn words of innocence and poverty. That Court had the benefit of live testimony where the demeanor of the witnesses would be gauged and evaluated. Our veto power on the Tax Court’s conclusion as to credibility is not of Niagara proportions.
Synopsizing, we find (1) persistent and large deficiencies, (2) admittedly false information given to bookkeepers, (3) unheeded warnings to keep adequate records, and (4) an agreement as to fraud in prior years. Tax evaders seldom leave tracks and therefore circumstances can be convincing. Though an isolated erroneous tax figure cannot be escalated or pyramided into fraud, a confraternity of similar errors can take on more sinister tax aspects.
We can indulge in no presumption to penalize even an errant taxpayer, but we do not believe that this tutelage interdicts us to forego ineluctable inferences and common sense. This is not a case of criminal tax fraud where we must be satisfied that guilt was shown “beyond a resonable doubt.” Helvering v. Mitchell, 1938, 303 U.S. 391, 397, 58 S.Ct. 630, 82 L.Ed. 917, 920-921; Henry v. C.I.R., 5 Cir. 1966, 362 F.2d 640, 643. We assume that common sense can be the possession of the unsophisticated, the unschooled, and the unlettered. Webb, though not privileged to have the other ingredients, was in possession of common sense. He must have, by the exercise of common sense, known that the figures used in his tax return were serious understatements of his obligations to his government. Innocent falsity is generally improbable; here it is impossible.
Affirmed.