Spatio-Temporal Analysis of Regional Unemployment Rates
A Comparison of Model Based Approaches
DOI:
https://doi.org/10.57805/revstat.v16i4.255Keywords:
unemployment estimation, model based methods, Bayesian hierarchical modelsAbstract
This study aims to analyze the methodologies that can be used to estimate the total number of unemployed, as well as the unemployment rates for 28 regions of Portugal, designated as NUTS III regions, using model based approaches as compared to the direct estimation methods currently employed by INE (National Statistical Institute of Portugal). Model based methods, often known as small area estimation methods (Rao, 2003), “borrow strength” from neighbouring regions and in doing so, aim to compensate for the small sample sizes often observed in these areas. Consequently, it is generally accepted that model based methods tend to produce estimates which have lesser variation. Other benefit in employing model based methods is the possibility of including auxiliary information in the form of variables of interest and latent random structures. This study focuses on the application of Bayesian hierarchical models to the Portuguese Labor Force Survey data from the 1st quarter of 2011 to the 4th quarter of 2013. Three different data modeling strategies are considered and compared: Modeling of the total unemployed through Poisson, Binomial and Negative Binomial models; modeling of rates using a Beta model; and modeling of the three states of the labor market (employed, unemployed and inactive) by a Multinomial model. The implementation of these models is based on the Integrated Nested Laplace Approximation (INLA) approach, except for the Multinomial model which is implemented based on the method of Monte Carlo Markov Chain (MCMC). Finally, a comparison of the performance of these models, as well as the comparison of the results with those obtained by direct estimation methods at NUTS III level are given.
Downloads
Published
How to Cite
Issue
Section
License
Copyright (c) 2018 REVSTAT-Statistical Journal
This work is licensed under a Creative Commons Attribution 4.0 International License.