Content area
Full text
Key points
* Respondents to the LFS are interviewed up to five times, at intervals of three months.
* The LFS longitudinal datasets link together the data from these quarterly interviews, enabling analysis of the changes experienced by individuals over time.
* Linking successive quarters shows that more than 95 per cent of those in employment at the first quarter remain in employment three months later.
* Of those unemployed at the first quarter, only around a half remain unemployed three months later.
* Roughly the same number of people flow into employment from economic inactivity as do so from unemployment.
Linking together data on the same individuals across two quarters of the Labour Force Survey can provide a source of information on the effects of labour market policies. This is the first of a series of articles that will analyse the longitudinal datasets in more detail
Introduction
SINCE 1992 the Labour Force Survey (LFS) has been conducted on a quarterly basis with a panel design. The sample contains around 60,000 households, and each household is interviewed five times at quarterly intervals, with a fifth of the sample being replaced each quarter. Although the survey is designed to produce cross-sectional data, there is also the potential to link together individuals' responses to provide longitudinal data, which can be used for a whole new range of analyses of the changes people experience over time. ONS has recently made publicly available two sets of LFS longitudinal datasets, linking respectively two and five consecutive waves of data, and including all people of working age who responded at each of the waves.
One of the most important uses of the new datasets is to enable analyses of movements between the three main economic activity categories of employment, unemployment and economic inactivity. Cross-sectional estimates of the stocks of the employed, unemployed and inactive only show the net flows between these three categories. However, with the longitudinal data it is possible to look at the full picture of the gross flows into and out of each of the categories. For example, while the cross-sectional data show that the number in employment fell by 76,000 between summer and autumn 2000 (working age, not seasonally adjusted), the longitudinal data...