Author:
Mushkudiani Nino,Pannekoek Jeroen
Abstract
In this paper we investigate the application of macro-integration methods to combine two sources of labor force statistics: a survey and an administrative source. In particular, we aim to arrive at a single estimate of the time series of temporary employment that efficiently combines the information from both sources. By varying the specifications of the objective function and constraints, four different macro-integration models were defined. The most plausible results were of a model that treats neither of the sources as fixed and uses multiplicative adjustments. The results were compared with the previous research where a latent Markov model was used to estimate the same time series. This Markov model approach does not lead to very different estimates of the time-series of temporary (or permanent) employment contracts but results in smaller estimates of the proportion of “movers”, persons that change contract status from temporary to permanent or the other way around. The model-based approach also provides estimates of the measurement errors in each of the sources. On the other hand, the macro-integration approach is less restrictive in the sense that it does not impose a Markov property of the integrated times series of proportions and it is more easy to implement.
Subject
Statistics, Probability and Uncertainty,Economics and Econometrics,Management Information Systems
Reference19 articles.
1. Bakker BFM. Micro integration: State of the art. ESSnet on Data Integration Report of WP1. State of the art on statistical methodologies for data integration, Eurostat; 2011.
2. Topics of statistical theory for register-based statistics and data integration;Zhang;Statistica Neerlandica,2012
3. Fosen J, Zhang LC. Quality evaluation of employment status in register-based census. In: Int Statistical Inst: Proc 58th World Statistical Congress, 2011, Dublin (Session STS024); 2011. Available from: http://2011.isiproceedings.org/papers/650264.pdf.
4. Guarnera U, Varriale R. Estimation from Contaminated Multi-Source Data Based on Latent Class Models. ISTAT; 2015. Available from https://ec.europa.eu/eurostat/cros/system/files/Guarnera-etal_abstract_NTTS_2015_Guarnera_Varriale.pdf.
5. de Waal T, van Delden A, Scholtus S. Multi-source Statistics: Basic Situations and Methods. International Statistical Review. 2019; Accepted for publication.