ABSTRACT. As pointed out by the organisers of this conference, sequence analysis is essentially an exploratory tool and needs to be complemented with other modelling approaches when it comes to testing hypotheses or studying the dynamics that drives the trajectories. This paper will first explore some of these tools: event-duration models which lead to event history analysis; event-sequences models which lead to sequence analysis; multiple level models which lead to multilevel analysis; social or geographical network models which lead to social-network analysis; models based on individual agents which lead to agent-based analysis. It then shows that these models can be classified under some more general paradigms: the event-history paradigm covers event history and sequence analysis; the multilevel paradigm covers multilevel and social-network analysis. Only the agent-based analysis seems to escape from these paradigms as it incorporates various social processes. However as it remains at the individual level it is too reductionist to explain social behavior. It seems then necessary to set up a more robust research program for demography. This research program may follow the induction’s way given by Bacon in searching for the structure of the studied phenomena and the interactions between the networks created by people. Such a program will be able to lead to a convergence of these different models.

ABSTRACT. Looking back at the progress made collectively since the introduction of sequence analysis (SA) in the social sciences enables, not only to feel good about the achievements, but also to understand better the dynamics of stability and innovation in the field. To this purpose, the communication will successively: analyse the specific conditions of development of SA; discuss issues present in reviews of literature and brief historical accounts of SA, as well as two contradictory debates, about the relevance of SA (SMR 2000) and about the best approach to multidimensional sequences (SM 2015); and code and process an extensive sample of sequence analyses so as to produce a map the field. This should enable to: 1. Detect trends and turning points; 2. Spot the factors of methodological change, both internal or external to the intellectual sphere; 3. Understand why some paths have been taken, and others sealed off; and 4. En-courage alternative options. (FULL ABSTRACT ATTACHED)

ABSTRACT. The paper analyses the daily activities of dual-earner couples in Italy. The goal is to discover how Italian dual-earner couples organize their daily activities (sleep, personal care, work, moving, housework, free time), during a typical work day from Monday to Friday. The analysis, carried out on data from the 2008 Italian Census on Time Use (the last one available), involves all the 873 couples that filled in their diaries on the same day.

Using the binary index (Bison, 2006, 2011a, 2011b, 2014; Bison, Rettore, Schizzerotto 2010; Franzosi, Bison 2010), we conduct a ‘multichannel’ analyses on the dual-earner couple’s activities from 7.00 to 22.00.

Quite strong relations with socio-demo-geographic conditions emerge from these analyses. Hence there is a strong relation of time packaging and the time spent on the various activities according to where couples live with respect to both geographical area (North, Centre and South & Islands) and the size of the town (metropolitan, more than 50,000 residents, from 10,000 to 50,000, fewer than 10,000).

Strong relations also emerge with the level of education, the social class and the occupational sector of Him and Her. Relations with the presence of children are observed mainly at the beginning and the end of the day. At the same time, the different time packaging profile of the dual-earner couple that emerges from the k-means cluster analysis seems to have a direct effect on His and Her level of satisfaction.

All the preliminary analyses seem to confirm the idea that dual-earner couples package their life time mainly in accordance with their jobs. Moreover, the analyses show that this time packaging changes in relation to the kind of job (social class) and the occupational sector. Secondly, the time spent on each activity changes according to the level of education of Him and Her, but there is an additional effect due to the social and cultural level of the area where they live.

ABSTRACT. Long working hours and high work load are possible risk factors for depressive symptoms, but relatively few prospective studies have been conducted thus far. Little is also known about the influence of the work environment on the course of depressive symptoms. This study aimed to use trajectories of depressive symptoms across working age to examine whether work load from paid and un-paid work predict these trajectories. The study was based on data from 4 waves of the Swedish Longitudinal Occupational Survey of Health (SLOSH, 2008-2014). We applied latent-class growth modelling in order to identify groups with similar development of depressive symptoms (depression trajectories) between ages 20 to 65. We further studied workload from paid and unpaid work as risk factors for certain pattern of symptoms and time-varying modifiers of trajectory level within each group. Six groups were identified with symptoms: ‘stable very low’, ‘stable low’, ‘mild decreasing’, ‘recurrent mild-moderate’, ‘stable moderate-high through midlife’, ‘stable high’. More hours in unpaid work in 2008 increased the risk of belonging to the groups with ‘stable moderate-high through midlife’ and ‘recurrent mild-moderate’depressive symptoms. When treating workload from unpaid and paid work as time-varying covariates we found that workload from unpaid work was associated with higher depressive symptoms although the effect was not statistically significant for all groups. On the other hand higher workload from paid work in terms of long working hours resulted in higher depres-sive symptoms for all groups except for individuals belonging to the group ‘stable moderate-high through midlife’. Females predicted depressive symptoms in all groups when compared to the ‘stable very low’ group. This study supported that are heterogeneous individual patterns of depressive symptoms over the working life. The results also indicate that workload both from paid and unpaid work influences the evolution of depressive symptoms across age.

ABSTRACT. Earlier research has shown that mental health problems at early stages of life negatively affect the individuals’ long-term employ-ment prospects and occupational attainments. Accordingly, the chances of individuals with mental disabilities of becoming economically independent are often constraint and their social integration is hampered. This situation impairs their well-being which in turn is likely to exacerbate their mental health. One of the reasons for the limited occupational achievements of individuals with mental disorders is their lack of educational de-grees. They may have dropped out of school, have achieved only low levels of education or have graduated from schools for children with special needs. Their educational attainments thus often do not allow them to take up standard jobs. Our paper addresses the question how mental health interventions in children with mental disabilities are linked to their labor market outcomes. We analyze whether there are educational and vocational programs or psychiatric interventions that mitigate the negative impact of mental disorders on educational and occupational outcomes. Moreover, we examine which consecution of educational and psychotherapeutic measures enhances young adults’ chances to be in employment at age 21. We analyze a dataset on 500 young adults in Switzerland who have either received a disability pension or rehabilitation measures between 2010 and 2013. We find that the earlier in life interventions – such as rehabilitation measures – are implemented, the higher is the chance that individuals are in an economic activity in early adulthood. Our analysis shows that individuals who entered disabil-ity insurance system relatively late and exhibit frequent interruptions in their educational trajectory are likely to be without training or employment at age 21. This finding is in line with the life course paradigm that suggests that individuals who differ with respect to a particular characteristic at a young age – for instance mental health – manifest much stronger differences in this characteristic later in life. Moreover, focusing on rehabilitation measures, we find that a large share of individuals has experienced continuous psychiatric treatment. The implications of our findings are that there is a need for early intervention, particularly in school settings for children with mental disorders. The earlier the intervention takes place, the better there seems to be the outcome for the affected individuals.

ABSTRACT. This study focuses on young adults with disabilities and their pathways towards work and family in past society. The aim is to explore their life trajecto-ries and compare them to a non-disabled group of people who experienced the same time-space context, represented by the 19th-century Sundsvall region, Swe-den. We employ sequence analyses on a series of demographic events that were to occur in the life of young adults: first occupation, marriage and parenthood. We also check for the events of death and out-migration. Disability studies show that disabled people were often subject to stigmatization caused by their impairment and prevailing perceptions about normalcy in in society. This would have limited their opportunities of work and family compared to non-disabled persons. Individ-ual-level data consisting of parish registers digitized by the Demographic Data Base (DDB), Umeå University, Sweden, allow sequence analysis that helps to an-swer the questions of whether and how disability influenced people’s life trajecto-ries. We obtain a holistic picture of how their life developed that suggests that dis-ability substantially limited people’s opportunities to find job, marry and form a family. This indicates that a stigma was associated with disability beyond the im-pairment itself and worked to add to disabled individuals’ difficulties in both the labor market and marriage market.

ABSTRACT. In this article, we consider the debate on the effects of the increasing flexibility in employment relationships on employment security, extending it to the long-term perspective of career studies. We explore life-long career sequence patterns in 13 European countries, with two objectives in mind. First, we examine to which degree the empirical career patterns correspond to career types postulated in the respective literature. Second, we consider how employment security has evolved over time, depending on the empirical career type. Our analyses indicate that some of the highly mobile career patterns are characterized by a higher degree of precariousness, a dynamic generally overlooked in mainstream career studies. In terms of policy, our results suggest the need to combine labor market flexibilization with individual supportive measures for some workers.

ABSTRACT. The aim of this contribution is to understand to what extent activation policies are able to reduce the risk of transition into the labour market of young people unemployed or in social assistance. The relevance of the institutional structure in which activation policies are implemented and the importance of the nature of activation policies on young people’s transitions are assessed by using optimal matching and event history analysis. The theoretical background underpinning the analysis is an original combination of the Capability Approach (CA)(Sen and Nussbaum 1993) and Transitional Labour Markets Approach (TLM)(Schmid and Gazier 2002; Schmid 2008). These theoretical frameworks, together, are believed to provide new perspectives on the assessment of activation policies. Their combination allows the researcher to assess whether the nature of activation policies shows different capacity of compensating for the lack of ‘transitional skills’ of the most disadvantaged young people (Bijwaard and Veenman 2008). The analysis is performed on a customized longitudinal administrative dataset and it focuses on the case study of the Brussels Capital Region (RBC).

ABSTRACT. The submitted extended abstract presents an analysis of administrative records from Luxembourg whereby the work-family trajectories of parents are re-constructed and analyzed using tools from state sequence analysis and a clustering algorithm is applied to search for typical patterns. In the final step available co-variates are linked to cluster membership via a multinomial logit model.

ABSTRACT. Within the social and economic sciences and of particular interest to demographers are life course events. Looking at life sequences we can better understand which states, or life events, precede or are precursors to vulnerability. A tool that has been used for policy evaluation and recently has been gaining ground in life course sequence simulation is dynamic microsimulation. Within this context dynamic microsimulation consists in generating entire life courses from the observation of portions of the trajectories of individuals of different ages.

In this work, we aim to use dynamic microsimulation in order to analyse individual professional trajectories with a focus on vulnerability. The primary goal of this analysis is to deepen upon current literature by providing insight from a longitudinal perspective on the signs of work instability and the process of precarity. The secondary goal of this work which is to show how, by using microsimulation, data collected for one purpose can be analysed under a different scope and used in a meaningful way.

The data to be used in this analysis are longitudinal and were collected by NCCR-LIVES IP207 under the supervision of Prof. Christian Maggiori and Dr. Gregoire Bollmann. Individuals aged 25 to 55 residing in the German-speaking and French-speaking regions of Switzerland were followed annually for four years. These individuals were questioned regarding, amongst their personal, professional and overall situations and well-being. At the end of the fourth wave, there were 1131 individuals who had participated in all waves. The sample remained representative of the Swiss population with women and the unemployed slightly over represented.

Using the information collected from these surveys, we use simulation to construct various longitudinal data modules where each data module represents a specific life domain. We postulate the relationship between these modules and layout a framework of estimation. Within certain data modules a set of equations are created to model the process therein. For every dynamic (time-variant) data module, such as the labour-market module, the transition probabilities between states (ex. labour market status) are estimated using a Markov model and then the possible outcomes are simulated.

The benefit of using dynamic microsimulation is that longitudinal sample observations instead of stylised profiles are used to model population dynamics. This is one of the main reasons large-scale dynamic microsimulation models are employed by many developed nations. There has been limited use, however, of such approaches with Swiss data. This work contributes to the analysis of professional trajectories of the active Swiss population by utilising dynamic microsimulation methods.

ABSTRACT. Recent theories about social and demographic change, such as individualization theory and Second Demographic Transition theory, suggest the emergence of a type of late, protracted and complex pathway to adulthood. In recent years, studies offer qualified support for the emergence of this new pattern of transition to adulthood in most European countries. However, the transition to adulthood is a complex process of a series of events that are often interlinked. Even though life courses are greatly varying sequences of roughly the same life course events, the complexity is caused by the fact that these sequences consist of correlated events and spells and these correlations depend on gender, social class, cohort and cohort-related macro events. Our previous work demonstrated that the application of stochastic models like the Latent-Class model helps to describe the variation in life courses and its correlation with gender and social class. But the Latent-Class model cannot account for correlated events within life courses nor can it account for switches between latent types during the life course. We argue that (Hidden) Markov models, as a simple generalization of the Latent-Class model, has the ability to account for correlations between events and spells and also allows for switches between latent types or “model life courses”. Therefore, this study will use (Hidden) Markov models to produce a typology of trajectories of the transition to adulthood. Furthermore, we will test hypotheses on social class- and gender differences in observed life courses and latent types or “model-life courses”, using data from the Gender and Generation Programme (GGP), which provides full monthly life course sequence data between age 15 to 40.

ABSTRACT. A hidden Markov clustering procedure is applied to a sample of n=185 longitudinal Internet Addiction Test trajectories collected in Switzerland. The best solution has 4 groups. This solution is related to the level of emotional wellbeing of the subjects, but no relation is observed with age, gender and BMI.

ABSTRACT. Panel data are valuable for answering questions about change over time, but remain relatively scarce in most developing countries, especially sub-Saharan Africa (SSA). Where there exists a series of repeated cross-sectional data, ‘pseudo-panels’ provide a promising alternative. The use of ‘pseudo panels’ has received considerable attention in econometrics, but application in Demography remains rare. This paper explores the potential for using ‘pseudo panels’ to investigate causal link between HIV and fertility in SSA. The relationship between HIV and fertility is a complex one, partly because causality can run in either direction. We focus primarily on fertility as the outcomes of interest and HIV as a contributing factor. Repeated cross-sectional Demographic and Health survey (DHS) data from 20 countries in SSA are used to construct “pseudo panels” based on birth cohorts by country. The pseudo panels allow an investigation of possible causal link between HIV in an earlier survey and fertility behaviour of similar cohorts in a subsequent survey. Measures of HIV are based on risk perception and HIV status, while fertility is based on births within the last five years and future fertility intention. A total of 140 cohorts (7 age groups * 20 countries) were constructed, with 120 cohorts having data for at least two time periods. The analysis used two alternative approaches: (i) ‘Conditional’ models of HIV in an earlier survey and fertility behaviour in subsequent surveys; and (ii) Repeated measures multilevel analysis, with cohort as Level-2, and measurement occasion as level-1. An evaluation /assessment of the analysis involved a comparison of findings from ‘pseudo cohort’ with individual-level analysis, and a multi-level estimation of intra-cohort correlation coefficients to assess the degree of similarity of individuals in the same cohort. A multivariate analysis based on fixed effects models was used to determine the extent to which observed patterns may be attributable to key demographic/ socio-economic differences or infer possible causal links. Preliminary analysis shows promising results on application of ‘pseudo panels’ in investigation of demographic causal links in settings with limited panel data such as sub-Saharan Africa. However, further analysis is necessary for conclusive results. In particular, advanced modelling using Multiprocess modelling or Structural equation modelling will be used to address possible endogeneity in the relationships observed.

ABSTRACT. Using data from the German Socio-Economic Panel on children born between 2002 and 2006, this study investigates the relationship between mothers’ monthly employment sequences in the first five years after birth and their children’s body weight around age six. Employment history is measured by different variables derived from sequence analysis that each capture specific aspects such as typical employment patterns (based on additional cluster analysis), complexity, and turbulence. The association between these measures and children’s body weight is estimated by inverse probability of treatment weighting of marginal structural models, a method developed in epidemiology to deal with the challenges of estimating the effects of time-varying exposures such as maternal employment. The results indicate that children who experienced very different maternal employment sequences but are similar with regard to background characteristics such as maternal education, household income, and family structure show no substantively or statistically significant disparities on body weight around age six. The study goes beyond the literature by focusing on maternal employment history beyond simple duration measures and by explicitly accounting for time-varying confounders of the relationship between maternal employment and child development.

ABSTRACT. In France there is a lack of accurate and up-to-date data on care pathways and care consumption of patients affected with multiple sclerosis (MS). The aim of this study was to describe care consumption of MS people in France and build a typology of their care pathways over the 2007-2013 period. To answer this issue, sequence analysis and agglomerative hierarchical clustering were used on a random sample of 1,000 patients, issued from French health-care databases, split according to the data available. The typologies were then described according to individual characteristics and care consumption. Two similar partitions, using indel costs fixed to 0.9 and transition-based substitution costs, were obtained : a five-cluster and a six-cluster typology for respectively complete and incomplete care pathways (i.e. less than seven years). Low care consumption was associated with older and less treated patients. Patients having a medium-low care consumption were younger and treated, whereas those having a high consumption seems to be older and have more comorbidities than the others. This pioneer study, using an innovative method in health field, gives a first overview of the care consumption of MS-affected people in France using objective and quantitative information. To go further, same work will be performed on the whole French population affected with MS and will include biological and medical imaging exams, and specialists’ care to complete care pathways.

ABSTRACT. When analysing complex sequence data with multiple channels (dimensions) and long observation sequences, describing and visualizing the data can be a challenge. Hidden Markov models (HMMs) and their mixtures (MHMMs) offer a probabilistic model-based framework where the information in such data can be compressed into hidden states (general life stages) and clusters (general patterns in life courses).

We studied two different approaches to analysing clustered life sequence data with sequence analysis (SA) and hidden Markov modelling. In the first approach we used SA clusters as fixed and estimated HMMs separately for each group. In the second approach we treated SA clusters as suggestive and used them as a starting point for the estimation of MHMMs.

Even though the MHMM approach has advantages, we found it to be unfeasible in this type of complex setting. Instead, using separate HMMs for SA clusters was useful for finding and describing patterns in life courses.

ABSTRACT. This is an introduction on discrete-time Hidden Markov models (HMM) for longitudinal data analysis in population and life course studies. HMM belong to the class of Markov based models. In a Markovian perspective, life trajectories are seen as the result of a stochastic process in which the probability of occurrence of a particular state or event depends on its history. Markovian models are used to analyze the transition process between successive states. Unlike a conventional Markov process where each state is liked to the next one, in the hidden Markov model the current response is driven by a latent variable that follows a Markov process. The paper presents also a simple way of handling categorical covariates to capture the effect of external factors on the transition probabilities. Empirical illustrations using data on self reported health demonstrate the relevance of the different extensions for life course analysis.

ABSTRACT. This paper demonstrates how Russian men and women from different generations undertake their first demographic and socio-economic events. For the first attempt at understanding how these events are ordered, we used the descriptive techniques of Sequence Analysis, including chronograms, parallel coordinate plots, the durations of statuses, and the frequencies of subsequences on tables. The study was performed on a panel of the Russian part of the “Generations and Gender Survey” (GGS: 2004, 2007, and 2011). The subsample consists of 4,857 respondents of 1935-1984 years of birth. Our analysis reveals the changes between sexes and generations. Men devote a significant part of their youth to achieving socio-economic events, while women much earlier and more actively initiate their demographic careers. Nevertheless, by the age of 35, there are more respondents among men who have children and relationships than among women. Young people, compared to older generations, much more actively enter into cohabitations and have children in them, but they delay the onset of all these events, especially childbearing, to later ages.

ABSTRACT. How does the intergenerational transmission of disadvantage come about? This study aims to broaden our understanding by examining the extent to which income trajectories in later stages of young adulthood are influenced by the work- and family-related pathways young people take into adulthood. The transition to adulthood is a demographically dense period, in which individuals make important decisions regarding their future career and family life, which in turn are likely to have a large impact on their future earnings. This study assesses to what extent family background, in terms of parental income, education, family structure and race, is mediated by the career and demographic trajectories that youths choose during the transition to adulthood. It is examined to what extent incomes diverge between those opting for different pathways to adulthood and whether within groups choosing for the same pathway to adulthood, family background remains to have an influence on these income trajectories. This study uses panel data from the United States, the National Longitudinal Survey of Youths of 1997 (N=4966). Sequence analysis is used to define different career (based on education and employment) and demographic trajectories (based on household, relationship and parenthood status) between the age 17 to 25 (monthly spells), separately for men and women. The family background variables and the different clusters are included as categorical variables in a growth curve model, with annual income between ages 25 to 32 as the dependent variable. Results indicate that the effects of family background variables mostly disappear once the career and demographic trajectories are included. Career trajectories appear to be more important in explaining differences in income trajectories in early adulthood than demographic trajectories. Incomes diverge for individuals who are in career clusters with longer college enrollment compared to those who are in clusters that have little college education.

ABSTRACT. Sequence analysis has relatively high demands regarding data properties: the in-formation to be analysed needs to be longitudinal, is not allowed to have any gaps, and should contain categorical information for the statuses, which is exhaustive and mutually exclusive. Such kind of information is provided by either retrospective life course surveys, or administrative register data, or panel datasets. These kinds of datasets have different advantages and disadvantages regarding sequence data. The validity of sequence information in a particular dataset depends to a large extent on the recall bias and how the survey mode moderates it. In retrospective data, people were surveyed on one time point and have to re-member events of their life course, which might be decades ago. Since this cir-cumstance, there is a large literature on recall error regarding employment careers (e.g. Dex & McCulloch, 1998; Horvath, 1982; Jacobs, 2002). However, whether there is an effect of recall bias on the properties of sequences generated from ret-rospective data is not researched so far. Administrative data (e.g. on employment histories) are continuously collected, even if the intervals are varying. In general, they are collected by institutions or third persons, so that recall bias doesn’t play a role. Of course, administrative data have other disadvantages – such as limited number of variables available or distortion by the data generation process. How-ever, a comparsion of administrative and retrospective data of the same individu-als could help to find out, how recall bias influence sequence data. Furthermore, panel data, where people are surveyed repeatedly, collect the lon-gitudinal information on fixed time points, generally each year. Typical represent-atives are household panels, such as the German Socio-Economic Panel (SOEP) or the British Household Panel Study (BHPS). The recall bias in these datasets is comparatively low, because the distance between the event and the survey of the event remains small, i.e. people have to take information from their memory after 1.5 years on average. Although the expected recall bias should be the same be-tween different yearly surveyed household panels, we find remarkable differences between the properties of sequences between the SOEP and the BHPS (see figure). In order to assess this problem, I have created comparable samples of 16- to 18-year old individuals from both datasets and look at their employment statuses for five years using only very basic statuses, i.e. employment, unemployment, inactivity and education. First of all, there are structural differences as one would expect: In the UK, young people enter the labour market earlier than in Germany, where young people stay longer in education. This explains the higher share of employment (green) in the UK and the higher share of education (blue) in Germany. The second thing that is eye-catching, are the regular ‘waves’ in the SOEP data. These regularity doesn’t have its origin in the real phenomenon of employment statuses, but in the survey mode applied: in the SOEP, employment careers are surveyed as calendar data, while in the BHPS there are surveyed in an episode format, where in each year there are questions regarding the beginning and the end date of an episode. In the former instance it seems that respondents simply make their crosses for every twelve months, whereas this ‘wave’ structure seems to be much flatter in the BHPS data. However, the survey mode seems to influence crucially the properties of the data. The paper proposed aims at clarifying two issues: First, I want to estimate the recall bias of retrospective life course data compared to register data. Here, the data from the National Education Panel Study (NEPS, starting cohort 6) provides the unique possibility to combine the same individuals from the retrospective survey data with administrative data from the German employment service provided by the Institute for Employment Research (IAB). This allows assessing exactly the influence of recall bias on sequence characteristics. Second, with the comparison of the two datasets mentioned above (BHPS and SOEP), I would like to find out more about the effects of different survey modes on sequence characteristics, such as turbulence, episode number etc.

ABSTRACT. The aim of this study was to perform a diachronic analysis, using three complementary techniques, of behaviors in fencing: lag sequential analysis, polar coordinate analysis, and T-pattern detection. The calculations for the methods employed in this study were made using the following software programs, which are all freely available: GSEQ5 (lag sequential analysis) (Bakeman and Quera, 2011), HOISAN v.1.6.3. (lag sequential analysis and polar coordinate analysis) (Hernández-Mendo, López-López, Castellano, Morales-Sánchez and Pastrana, 2012), and THEME v.6 Edu (T-Pattern detection) (Magnusson, 1996, 2000).

We attach a extended abstract.