Skip to content

LFS Single Month Analysis (Not designated National Statistics): September 2011 This product is designated as National Statistics

Released: 14 September 2011 Download PDF

Abstract

Movements in the LFS data series at the end of 2003 prompted ONS to conduct detailed analysis of the LFS data to determine the reasons behind these movements. Experimental analysis of the data, at the highest aggregate level, was carried out to break the LFS data down into single month periods. This analysis proved useful so has since been produced every month.

Single Month LFS Analysis

Background

Movements in the LFS data series at the end of 2003 prompted ONS to conduct detailed analysis of the LFS data to determine the reasons behind these movements. Experimental analysis of the data, at the highest aggregate level, was carried out to break the LFS data down into single month periods. This analysis proved useful so has since been produced every month.

As the 3-month average of the single month series tracks the changes seen in the published LFS, it is possible to consider changes in the published LFS in terms of movements in the single month series. This in turn makes it easier to determine whether the movements in the published LFS series are true reflections of changes in the wider economy, or whether they are movements that reflect the survey nature of the LFS and its sensitivity to factors such as sampling error.

Method

Briefly, single month LFS estimates were produced by taking the raw, or unweighted, LFS survey responses for each month and weighting them up to single month population estimates, using a simplified weighting method (broad age band and sex). These single month estimates were then seasonally adjusted. By constructing a 3-month average of the seasonally adjusted single month series and comparing this with the published LFS (itself a 3-month average), it was possible to show that the average of the experimental single month series tracked the changes in the published series. The two series were not identical, however, due to the relative crudity of the weighting method used to produce the single month estimates. This tended to slightly overemphasise the weighting of those with higher employment rates. To remove the differences between the single month and published LFS, the single month series was benchmarked to the published LFS series, using a set of iterative equations. Further details of the method are included in the technical note at the end of this paper.

Charts

The charts in this briefing show the published and single month estimates for employment, unemployment and inactivity (seasonally adjusted). For the published series, the dates shown relate to the last month of the three (e.g. January - March is shown as March).

Employment Rates (16-64) Seasonally Adjusted

The figure for July 2011 shows a decrease of 0.5 percentage points on the previous month.

UK 16-64 Employment Rates (Seasonally Adjusted)

Employment rates for 16-64 over a 3 year period, September 2011
Source: Labour Force Survey - Office for National Statistics

Download chart

 

UK 16+ Unemployment Rates (Seasonally Adjusted)

The figure for July 2011 shows an increase of 0.2 percentage points on the previous month.

UK 16+ Unemployment Rates (Seasonally Adjusted)

Unemployment rates over the 3 years, September 2011
Source: Labour Force Survey - Office for National Statistics

Download chart

UK 16-64 Inactivity Rates (Seasonally Adjusted)

The figure for July 2011 shows an increase of 0.5 percentage points on the previous month.

UK 16-64 Inactivity Rates (Seasonally Adjusted)

Inactivity rates for 16-64 over 3 years, September 2011
Source: Labour Force Survey - Office for National Statistics

Download chart

Methodological article

A methodological article about these estimates is available on the website.

Technical Note

1. These single month LFS analyses have been produced as a tool to assist in understanding the movements in the published 3 month average LFS estimates. These estimates do not have National Statistics status and as such they are not suitable to be used as labour market indicators in their own right. It must be noted that they are based on only 1/3 of the usual LFS sample, and so are much more prone to the effects of sampling error.

2. Unweighted LFS survey estimates of main economic status, by broad age bands, sex and region were aggregated for each week from the microdata files. These data by week were then aggregated to form 4- or 5-week months based on the survey calendar.

3. For the single month estimates adjustment factors were calculated for, and applied to, the unweighted LFS aggregate data. This represents a simplification of one of the major stages of the LFS weighting method.

4. The single month estimates were then seasonally adjusted (SA) using X11Arima, using the same parameters as those used for the standard LFS estimates. To ensure additivity across age, economic status and region, the single month estimates were constrained, using the standard (top down) LFS constraining methods.

5. As described above, the averages of the single month series and the published LFS series were not identical, due to differences in the weighting regime used. The difference has been almost entirely removed through benchmarking the experimental series to the published LFS series. This was done by applying the differences between the published series and the average of the single month series to each single month point contributing to any given average. Although this method caused the differences between the two series to approach zero over time, multiple iterations (currently 40) are needed to reduce the differences below a significant level.

6. The method used to benchmark the single month includes an element of forecasting. As the accuracy of the latest data points increases with more data available the final three data points are subject to revision each month, but these revisions are all very small.

Background notes

  1. Details of the policy governing the release of new data are available by visiting www.statisticsauthority.gov.uk/assessment/code-of-practice/index.html or from the Media Relations Office email: media.relations@ons.gsi.gov.uk

Get all the tables for this publication in the data section of this publication .
Content from the Office for National Statistics.
© Crown Copyright applies unless otherwise stated.