Development and application of bayesian dynamic compartment modelling to study labour (in)stability in spain

  1. Carbonell Asins, Juan Antonio
Supervised by:
  1. Carles Simo Noguera Director
  2. Purificación Galindo-Villardón Co-director

Defence university: Universitat de València

Fecha de defensa: 29 July 2022

  1. Jordi Giner Monfort Chair
  2. Lucila Finkel Morgenstern Secretary
  3. Fernando Casanoves Committee member

Type: Thesis


This is a methodological thesis in which we focus on labour market. However, the methods here developed could also be applied to other areas of Social Science enabling researchers broadening knowledge by tackling research questions from a different point of view. In this context, Labour (in)stability could be operatized as transition speed from employment to non-employment and vice versa. If speed is high, instability is also high; on the contrary, if transition speed is low for a given set of individuals, these have low instability. This methodology might enable researchers identify new patterns in the study of labour precariousness that have yet not described. The primary aim of this thesis is to develop a novel methodological framework for the study of labour (in)stability which focuses on transition speed from one state to another. The data is an anonymized sample of one million people who were enrolled in Social Security in a specific year. Using this source, (In)stability will be studied from two perspectives. Firstly, as transitions from employment to non-employment and vice-versa in a two-compartment model. We will study market labour entrance cohorts, quinquennially, from 1970 to 1990 and we will be following them until 31/12/2013 as well as comparing by sex. Secondly transitions between fixed-time, permanent workers and non-employment in a three-compartment model. Here, due to sample limitations, individuals will be followed annually from 1991 to 2002 with a fixed follow-up of ten years and results will also be presented by sex. The conclusions of applying dynamic compartment models to study labour market for females and males goes in line to what other authors have already observed: the presence of strong gender labour segregation. Although we point out that the gap in transition speed between females and males to non-employment is narrowing, (in)stability patterns differ substantially between them.