A Representative Matched Cross-section Survey for Austria - Measuring Worker Flow Dynamics with the Austrian Labour Force Survey

While worker flow analysis has grown in importance in many countries, Austria still lacks a specific longitudinal dataset as a prerequisite to perform similar analyses. For this reason, this article provides a coherent procedure to construct a longitudinal dataset based on the rotational panel structure of the Austrian quarterly LFS from 2004 to 2014. The procedure, which is available for researchers, is grounded on the discussion of several related and important issues inherent in constructing this sort of longitudinal data: First, it deals with the construction of the quarterly matched dataset and the quality-of-measurement of several labour market variables. Second, the paper analyses non-response as a sample selection process, and shows that the selected (quarterly-matched) dataset causes biased estimates of worker flows. Third, the article proposes an iterative raking procedure to obtain survey weights as a bias-correcting device for any future analysis. Based on these adjustments, we present unbiased time-series of worker flows and transition rates, and conclude that the employment-unemployment margin is highly sensitive to economic shocks and that the Austrian labour market is additionally shaped by large movements within the participation margin.