LONGITUDINAL ANALYSIS METHODOLOGIES TO DETERMINE PROFILES OF INFORMAL WORKERS

IGLESIAS, MAXIMILIANO LUJÁN; STIMOLO, MARIA INES

Bahía Blanca

Congreso; XV Congreso Dr. Antonio Monteiro; 2019

Universidad Nacional del Sur (UNS) - INMABB CONICET

La publicación en Actas esta sujeta a referato y fue aprobada por los evaludores. El contenido de las actas es indizado por Mathematical Reviews.ABSTRACTOne of the main limitations for the correct analysis of the labor market in developing countries from a temporal variability approach is the lack of appropriate panel data information (Canavire-Bacarreza, Urrego, & Saavedra, 2017). The purpose of this article is to develop statistical methodologies that make it possible to incorporate the temporal dimension as a key factor to carry out the most complete possible analysis of the dynamics and structure of the problem under study, as well as the relationship between its multiple factors and determinants.The pseudo panel approach allows overcoming the limitation of data availability through the building of ?synthetic? panels and measuring the temporal evolution of characteristics of interest in cohorts of individuals and the building of ?variable-trajectories?.The temporal clustering approach, based on the non-parametric k-means algorithm, combines content similarities and temporal adjacency in a single representation which makes it possible to find cohort groups of homogeneous individuals in relation to the joint trajectories of their characteristics.With the results obtained according to the exploratory analysis, we could identify three well-defined groups based on the temporal trajectories of their informality rates and income levels.