Models of Multiple Events in the Analysis of Subsequent Registrations in the Labour Office

Abstract

In many fields of science, it is necessary to analyse recurrent events. In medical science, the problem is to assess the risk of chronic disease recurrence. In economic and social sciences, it is possible to analyse the time of entering and leaving the sphere of poverty, the time of subsequent guarantee or insurance claims, as well as the time of subsequent periods of unemployment. In these studies, there are different ways of defining risk intervals, i.e. the time frame over which an event is at risk (or likely to occur) for an entity. Research on registered unemployment in Poland shows a high percentage of people returning to the labour office and registering again. The aim of the article is assessment of the risk of subsequent registrations in the labour office depending on selected characteristics of the unemployed: gender, age, education, and seniority. In the study, methods of survival analysis were used. The results obtained for four models being an extension of the Cox proportional hazard model were compared. The Anderson‑Gil model does not distinguish between first and next events. The number of events that occurred is important. Two Prentince‑Williams‑Peterson conditional models and the Wei, Lin and Weissfeld models are based on the Cox stratified model. The strata are consecutive events. They differ in the way risk intervals are determined. In the analysed period, only age and education influenced the risk of multiple registrations at the Poviat Labour Office in Szczecin. Gender and seniority did not have a significant impact on this risk. The analysis performed for subsequent registrations confirmed the impact of the same features on the first subsequent registration. In general, it can be stated that the analysed characteristics of the unemployed did not have a significant impact on the second and subsequent returns to the labour office.

Keywords

• cox regression model
• risk of registration in the labour office
• recurrent events models