The body in charge of preparing official Government statistics was forced into an embarassing apology today after it completely miscalculated its estimates about the level of peoples' income at retirement.

The mistake by National Statistics means that its figures for how much pensioners were receiving from their private sector pension schemes could have undershot the real amount by as much as four times.

The error forced National Statistics to withdraw as inaccurate official figures that it published only five weeks ago.

It said it was checking to see whether its blunder had had any impact on estimates of the level of state pension provision.

"The estimates are being re-calculated and will be published again as soon as possible, once they have been quality assured," National Statistics said.

It could not provide a date for re-publishing its estimates. When they were first published, the apparently very low figures for private pension levels prompted a rash of headlines on pensioner poverty.

The admission is hugely embarassing for National Statistics, formerly known as the Office for National Statistics.

The establishing of National Statistics, an independent body free from government interference, was enthusiastically supported by Gordon Brown, the prime minister.

Mr Brown equated the importance of creating an independent statistics body with his decision to grant the Bank of England control over interest rates.

Estimates were first published on April 14 in a release entitled "Pensions provide only modest incomes for many".

Using data provided by the Department for Work and Pensions (DWP), it stated, incorrectly, that the average annual private pension income for retired couples in 2005/06 was £2,115.

But National Statistics said today that there had been a "programming error"

"The basic data used gave pension incomes as weekly amounts. In deriving annual estimates, some of these were treated as if they were monthly amounts," National Statistics said.

"Where this is the case, the correct annual estimates would be over four times as large as the figures published in April."

It said it could not be sure that the true figure for its example was four times the amount stated.

A spokesman said the error had been in the way that the data had been manipulated, not in the data itself. He said the error would have been spotted before it would have been used by other government departments.

Explore Economics

• Pre-Budget Report 2009

• Budget 2009

• World Economic Forum

• The Times MPC

Times Recommends

• Lies and cover-up at MG Rover exposed by report
• High oil prices push manufacturing costs up
• Primark founder steps down as chief executive