My right honourable friend the Secretary of State for Work and Pensions (John Hutton) has made the following Statement.

Households Below Average Income 1994-95 to 2005-06 was published on 27 March 2007. After publication the production team found that an error had occurred because incorrect population estimates were used to generate the dataset for 2005-06, upon which the publication was based.

Most of the key headline figures from the publication are unaffected by this error. Looking at the measures of relative poverty, correcting the error has had no impact on reported child poverty for 2005-06. However, the revised figures do see an increase in reported poverty among working-age adults, which is seen on both the “Before Housing Costs” and “After Housing Costs” measures. The revisions see both measures increase by an extra 100,000 under the rounding conventions, from 5.3 million to 5.4 million on the former basis and from 7.1 million to 7.2 million on the latter basis. Furthermore, although the revisions have no impact on the reported level of relative pensioner poverty, the rounding conventions we use now mean that we no longer see a change in measured pensioner poverty between 2004-05 and 2005-06.

Looking at measures of absolute poverty, the correction of the error has similar effects as on the measures of relative poverty, with no increase in reported measures of child poverty or pensioner poverty and a small increase in reported poverty among working-age adults.

Key headline results for the number of households in relative and absolute poverty, including changes as a result of the revisions, are being placed in the House Libraries. These revisions have been quality assured and verified by the Institute for Fiscal Studies.

The error does require reproduction of all the statistics and the re-release of the report. The changes to the HBAI statistics also affect other DWP publications, specifically Pension Credit Estimates of Take-up 2005-06 and The Pensioners’ Income Series 2005-06, but both to a minor extent, either in terms of the number of tables affected or the extent of the change.

We anticipate the full set of statistics being completed by the first half of May. We will announce a release date through the national statistical website, in line with National Statistics protocols.