Identifying Labour Market Dynamics Using Labour Force Survey Data

Despite the increasing focus on labour markets dynamics, little is known about either the quality of the underlying data or the appropriateness of the standard methodologies frequently used to analyse labour market dynamics in Europe (the retrospective and matched files approaches). This paper is a first attempt to fill these gaps. In section one we highlight the problems associated with the use of the alternative methodologies. In section two, we provide evidence as to the size and nature of the errors arising from the use of these two approaches. The results of our survey validation work suggest that, due to problems of recall error and heterogeneous survey design, the retrospective approach tends to result in a considerable number of spurious transitions being recorded. Whilst the use of quasi-longitudinal LFS data should in theory overcome such problems, issues of sample attrition and, more importantly, misclassification error, tend to result once again in significant over-reporting of labour market transitions. For example, it is estimated that in the first quarter of 1994, 12.5% of the Spanish unemployed were incorrectly classified into a different labour market state. These results, whilst preliminary, emphasize the need to allow for the underlying error structure of LFS data in empirical studies of labour market dynamics, and draw into question the findings of existing work which fails to do so. We conclude by providing an illustrative example of one of the ways in which the error probability rates estimated in section 2.2 can be used to adjust the underlying data, so as to correct for potential errors arising from the use of the matched files approach. The results obtained suggest, however, that techniques developed in the US to correct the observed labour market flows for misclassification errors may not be readily applicable to Spanish data