The analysis of longitudinal studies

Recently, economists, sociologists and other social scientists have become increasingly interested in longitudinal studies and their analysis, in their attempts to understand the dynamics of economic and social processes. Previously, single time cross-sectional surveys and their analysis formed the primary basis for empirical investigations in the social sciences. Longitudinal studies, in which the same units are investigated on several occasions, over extensive periods of time, are expensive undertakings and are complex operationally and methodologically. Even when longitudinal data are theoretically available, e.g., from panel surveys, such as the Labour Force Survey (LFS) in the UK and the Current Population Survey in the US, technical and methodological problems considerably reduce their usefulness for longitudinal analysis. In addition, their overall time span for a single unit is relatively short (about two years) and their panel design is primarily geared to increasing the efficiency of cross-sectional estimates, rather than that of the analysis of gross flows or developments over time.