Predicting the stability of early employment with its timing and childhood social and health-related predictors: a mixture Markov model approach

Abstract

Background and objectives: To extend work careers, it is important to focus on all working-agedpeople, including young adults. However, there is little knowledge about the stability of early careersand its life course social and health-related determinants. The aim of this study is to identify whatkind of typical patterns of work participation and transitions between different statuses can beidentified among young adults after their first entry into the labour market, and to examine whetherthe timing of entry together with parental and own socioeconomic position and health predict earlywork participation patterns.Methods: We used the Finnish Birth Cohort 1987 including data from several registers from all59,476 children born in 1987 as well as their parents, followed until the end of 2015. We createdsequences of monthly work participation statuses starting from the calendar month when a cohortmember had their first at least six-month long spell in employment or entrepreneurship. Workparticipation statuses (8 categories) were collected for 60 months or until the end of the follow-up. Weincluded altogether 51,871 individuals with an entry into paid employment, who could be followedfor at least 12 subsequent months. Parental social and health-related covariates were included fromnational registers. We used first sequence analysis and cluster analysis to identify latent groups ofindividuals with similar work participation patterns. Next, we estimated a mixture Markov model thatallows for identification of latent classes of labour market attachment, estimation of labour markettransitions within classes, and prediction of class membership using childhood social and healthrelateddeterminants.Results: We observed that entry into the labour market as measured by six months in continuousemployment was not a permanent entry for many, not only due to unemployment and health-relatedexits but also due to studies and parental leave. During the first five years, we observed a fairly linearrelationship between the age at entry and early work participation pattern: individuals entering thelabour market at a later age were more likely to be in continuous employment thereafter (all elseequal). Regarding both own and parental social- and health-related childhood factors, latent classesshow clear patterns of accumulated advantage and disadvantage: more advantaged backgroundpredicted exits due to studies or – when following a late entry – stable employment, whiledisadvantaged background factors predicted more unstable work and long-term exits from the labourmarket.Conclusions: There are varied work participation patterns and transitions between employmentstatuses among young employees. More in-depth understanding of these patterns and theirdeterminants is important to plan targeted interventions and to promote more stable work participationamong young adults.

Sequence analysis: Its past, present, and future

“…Integrations of SA and Markov based approaches have also been proposed in the literature. For example, a two-step analysis combining SA and MM has been proposed as a workaround solution for the issue of complex models with an unknown model structure (Helske et al, 2018;Helske et al, 2021) and the Sequence Analysis Multistate Model has been used for modelling the relationship between time-varying covariates and trajectories of categorical states (Studer et al, 2018b).…”