Numbers at random

Press, Volume CXIV, Issue 33675, 26 October 1974, Page 1

Numbers at random

The Inland Revenue Department satisfied a reporter of “The Press” yesterday that numbers allocated to taxpayers accounts were selected at random. The department was at pains to establish that the methods by which the account numbers were allocated had no sinister overtones. A study of the allocation system was prompted by a string of letters to “The Press” from readers who had noted marked similarities between the numbers allocated to them and to their wives or other close relatives. Numbers allocated to salary and wage earners in Christchurch were issued by a computer in Wellington. The information given to the computer was minimal — the name and address of the taxpayer. This was taken from returns filed by salary and wage earners for the year ended March 31, 1974. The returns, in keeping with the department’s longstanding practice, were sorted into alphabetical order before processing to determine refunds. As this was done, the names and ad-

dresses of the taxpayers were printed on to sheets which were then sent to Wellington to be fed into the computer and numbers issued. The department denied that the numbers on the accounts contained built-in information of any sort. It was to be expected that many husbands and wives had similar numbers because the returns would have been filled in and filed about the same time. Several letter-writers said that their suspicions had been aroused by the similarity of numbers for themselves and their wives when they had been given an assurance in writing that the numbers had been selected at random. The department said that self-employed people at least could select their own numbers if they wanted to from a long list of numbers presented to them. A lecturer in mathematics at the University of Canterbury, Mr M. H. Smith, said in a letter that among the readership of “The Press” there would be a large number of tax-paying couples among whom it was quite reasonable to expect that one or perhaps several would receive similar account numbers.

A reader in electrical engineering at the university, Mr J. H. Andreae, said it would be suprising that a couple in Christchurch would receive a pair of numbers which could be treated as an indication of common information. “If five digits were used for coding personal information, the remaining three digits would be insufficient for coding identity,” he said.