In his letter (December 10), Anthony C. De Brincat claims that NSO said that its statistics on overtime earnings "had been understated for a number of years". Your correspondent is manifestly wrong in his claim. NSO made public an estimate of overtime earnings only once, by giving a range of Lm42 million to Lm47 million annually for the years 2004-2006. This estimate was communicated in letters to The Sunday Times and to another local newspaper on November 25.

NSO presumes that the issue of NSO revisions on this topic referred to by the correspondent was built up around the figure of Lm12 million in overtime earnings.

This statistic was not an estimate made by NSO. This office has unequivocally stated, and would like to reiterate, that this statistic did not emanate from its calculations.

The correspondent and other interested parties will appreciate that a fundamental component of any revision is a base figure.

Since the Lm12 million statistic relating to overtime did not emanate from NSO sources, then the whole issue of a revision effected by this office on this topic falls through.

This office accepts responsibility for the revisions it actually effects, primarily because legitimately-based and well-communicated revisions are a normal and accepted part of statistical work practices. But it will not own any revisions that are attributed to it without any authentic basis.