Labour is demanding an apology from Finance Minister Bill English, saying he knowingly relied on flawed data when making claims about inequality in New Zealand.
Treasury and Statistics New Zealand today admitted a major error in its calculations of household disposable income, which overestimated incomes among poorer households.
This meant the number of children living in poverty was underestimated by 20,000, while the level of income inequality was also underestimated.
English confirmed today that he had been told of the mistake before Christmas.
English has this year relied heavily on reports by the Ministry of Social Development's principal adviser on social sector strategy, Bryan Perry, which in turn relied on the Treasury data, to dismiss claims of growing inequality.
The admission from Treasury and Statistics NZ today did not change the trend of inequality, MSD said, but did increase it slightly.
Labour's finance spokesman David Parker said English should have admitted to the mistake immediately.
"It's outrageous that he has known since the OECD [Organisation for Economic Co-operation and Development] pointed out the error prior to Christmas that the data that the Perry report is based on is wrong," Parker said.
"If he's trying to say that [inequality] figures aren't getting worse when in actual fact there's been a $1.2 billion overestimate of income among low to middle income families, well, that's just misleading."
A statement from Treasury blamed the mistakes on human error and breakdowns in process in its relationship with Statistics New Zealand.
Treasury said the had no effect on the "real world".
"[T]here are no 'real world' impacts on New Zealanders from the miscalculations," Treasury chief economist Girol Karacaoglu said in a statement.
"There is zero effect on the individual or household benefit payments and tax credits people get, or the tax they pay."
The departments were "deeply disappointed" that they had not met expected standards, he added.
It appears that queries from the OECD about the figures on child poverty uncovered the mistake.
It stemmed from a simple case of double counting, with the accommodation supplement included twice – overestimating the impact on those who received it.
Statistics NZ said no staff had been disciplined as a result of the error.
Treasury said the staff involved no longer worked there.
The error has led to several reports being withdrawn from the Ministry of Social Development website because they relied on Treasury data.
MSD said the mistake meant that using the widely used – if controversial – measure of childhood poverty (those living in households with less than 60 per cent of median income after housing costs) in 2011/12 was 285,000, 20,000 more than previously calculated.
The number of children living in households with less than 50 per cent of median income was also higher than expected.
MSD also said that inequality, using the measure preferred by the Government, was also higher than previously estimated, albeit by a small amount.
"Inequality as measured by the Gini [coefficient] is slightly higher using the revised data, but the trend line is still flat and remains a little lower than the high point in 2004," MSD said.
English said the figures ‘‘reinforce the case for our strong focus on helping get people off welfare benefits and into work’’.
Should David Cunliffe lead Labour again?Related story: Cunliffe, Robertson trade blows