National Statistics Online - Articles - New Methodology for Low Pay Estimates
Articles Database
New Methodology for Low Pay Estimates
This article describes the new methodology for low pay estimates and provides estimates using the new methodology for 1998-2003.
A new methodology has been created for low pay estimates that uses data from the new Annual Survey of Hours and Earnings (ASHE), which replaces the New Earnings Survey (NES) as the first major survey to benefit from ONS' wide-ranging modernisation programme. The ASHE methodology includes imputation and a weighting methodology that is partly based on occupation, which is known to be a key variable in predicting pay, and therefore provides more accurate weights than those previously used for low pay estimates for the NES. In 2004, improved editing is being introduced, and supplementary surveys are included in the ASHE to improve coverage at the low end of the pay distribution, so that ASHE can appropriately be used as the sole basis for low pay estimates.
Apart from 1998, differences between the overall estimates using the previous central estimate methodology and the new ASHE methodology are small and within sampling error of each other. The fall in 1998 estimate from 1.4 million to 1.2 million jobs is due to the improvement in the weighting methodology. At the disaggregated level, results across 1998-2003 show: a slightly higher ASHE estimate for 18-21 year olds; a slightly lower ASHE estimate for those aged 22+, affecting men and women equally but part-time rather than full-time workers, particularly part-time men; a slightly higher estimate for full-time women; similar estimates for industry except for 1998 when improved weights reduce estimates for sectors H and O by 170,000 jobs; similar estimates for regions; and similar estimates for occupations except in 2003 for Occupation Group 9.
Published in web format: 15 October 2004 at 9:30 am
Download
