As the banking industry replaced manual bookkeeping with computerized systems as a means of keeping track of customer accounts, they unknowingly opened the door to a type of embezzlement that had previously been difficult (if not impossible) to successfully pull off on a large scale: the ‘salami’ technique. This was a scheme by which a bank employee (usually a computer programmer) could surreptitiously stockpile a substantial amount of money not by grabbing large sums all at once, but by ‘slicing’ off ‘thin’ amounts of cash from many different customers’ accounts and diverting them to one central account (which he controlled), thereby building up a tidy nest egg while minimizing the risk that any single instance of his theft would be detected and investigated.
Thomas Whiteside reported a supposedly real example of this technique in action over forty years ago, in his 1978 book
The embezzler was evidently using the bank’s computer to transfer twenty or thirty cents at a time, at random, from 300 checking accounts at the bank and diverting the money to a dummy account for his own use. The computer criminal was careful never to divert sums from any particular account more often than three times a year. Because a customer was unlikely to notice such a small discrepancy in his monthly bank statement — or, if he did notice it, to find it worth his while to go to the bank and argue over it — the embezzlement was likely to go on and on.
One way of taking the salami technique a step further was for the embezzler to cover his tracks by making even his small thefts of a few cents at a time look legitimate:
Two programmers who were employed by a big New York garment firm instructed the company’s computer to increase by two cents the amount withheld from their fellow-employees’ paychecks each week for federal taxes. They further programmed the computer to direct the two cents per employee per week to their own federal withholding accounts. The result was that at the end of the year they received the money in the form of refund checks from the Internal Revenue Service, which had been acting as an unwitting bagman for the embezzled sums.
Even better was a variation of this technique in which the vanished sums were so unnoticeably small that they could be uncovered only through the most rigorous of audits:
One way in which the computer criminals might employ the salami technique is to round down any sums ending in fractions to the nearest whole number — for example, fractions of pennies as these are computed in interest-bearing accounts. In the meantime, the criminal has established a dummy account at the same bank, and he programs the computer to divert the surplus from the round-downs to this account. Quietly accumulating year in and year out, these fractional sums can mount handsomely, and usually neither the bank nor the depositors know what is going on.
In classic urban legend fashion, one version has it that the unlucky thief who tried this last method fell victim to a fluke of a company promotion and was caught:
A programmer working at a mail-order sales company had its computer round down odd cents in the company’s sales-commission accounts and channel the round-downs into a dummy sales-commission account he had established under the name of Zwana. He had invented the name Zwana because he knew that the computer processed the company’s accounts in alphabetical order, and he could easily program the computer to transfer all the round-downs into the last account in the computing sequence. The system worked perfectly for three years, and then it failed — not because of a logical error on the culprit’s part, but because the company, as a public-relations exercise, decided to single out the holders of the first and last sales-commission accounts on its alphabetical list for ceremonial treatment. Thus Zwana was unmasked, and his creator fired.
Whiteside, Thomas. Computer Capers.
New York: Mentor, 1978. ISBN 0-451-62173-5 (pp. 33-35).