Abstract:
The Brazilian Labour Force Survey (BLFS) is a quarterly rotating panel survey with 80% sample overlap between two successive quarters. We will present time series models developed to produce model-based single month estimates at national level as well as small area (state-level) estimates, which are both at a higher frequency than those currently being published. In addition, multivariate time series models that integrate survey data and Google Trends series for nowcasting are considered. High dimensionality problems are solved using a dynamic state space model. In this case, we also discuss the choice of search terms and approaches for targeting predictors in the dimensionality reduction process. The models account for the autocorrelation due to sample overlap and the increased volatility in the labour force series in 2020. This is joint work with Luna Hidalgo (IBGE) e and Jan van den Brakel (Statistics Netherlands and Maastricht University).
25 January 2023
Location:
Watch this webinar