Holistic Trajectories: a Study of Combined Employment, Housing and Family Careers by using Multiple-Sequence Analysis

Abstract

Social science applications of sequence analysis have thus far involved the development of a typology on the basis of an analysis of one or two variables which have had a relatively low number of different states. There is a yet unexplored potential for sequence analysis to be applied to a greater number of variables and thereby a much larger state space. The development of a typology of employment experiences, for example, without reference to data on changes in housing, marital and family status is arguably inadequate. The paper demonstrates the use of sequence analysis in the examination of multivariable combinations of status as they change over time and shows that this method can provide insights that are difficult to achieve through other analytic methods. The data that are examined here provide support to intuitive understandings of clusters of common experiences which are both life course specific and related to socio-economic factors. Housing tenure is found to be of key importance in understanding the holistic trajectories that are examined. This suggests that life course trajectories are sharply differentiated by experience of social housing. Copyright 2007 Royal Statistical Society.

La mobilité comme variabilité temporelle de la présence spatiale

“…Figure 1) amène facilement à des analyses de présence spatiale ou de rythmes urbains telles que celles réalisées par Banos et Thévenin (2004) Finalement, le modèle de données séquentiel offre la possibilité de créer des séquences multi-couches (Pollock, 2007), c'est-à-dire des séquences qui combinent plusieurs couches d'états caractérisant les épisodes mobiles et immobiles. Un exemple simple serait de combiner les états désignant les activités (domicile, travail, achats, etc.)…”

Work and family over the life-course. A typology of French long-lasting couples using optimal matching

“…It was introduced into the field of social sciences by Andrew Abbott in the 1980s (Abbott & Forrest, 1986). Since then, optimal matching has been successfully applied to life courses and family histories (Billari, 2001;Schoon, McCullough, Joshi, Wiggins & Bynner, 2001;Aassve, Billari & Piccarreta, 2007;Pollock, 2007) and careers (Stovel, Savage & Bearman , 1996;Blair-Loy, 1999;Han & Moen, 1999;Robette & Thibault, 2008;AnyadikesDanes & McVicar, 2010), as well as many other topics such as couples' time-schedules (Lesnard, 2010). Despite a few criticisms (Wu, 2000;Elzinga, 2003), it has now become widely accepted as a useful tool for life course scholars (Aisenbrey & Fasang, 2010).…”

“…A first strategy involves first using optimal matching to compute one distance matrix for each dimension and then to carry out a linear combination of these matrices into one by means of linear combination (Han & Moen, 1999). A second strategy involves creating a new state variable combining the single states associated to each dimension (see for instance Blair-Loy, 1999;Aassve, Billari & Piccarreta, 2007;Pollock, 2007;Robette, 2010): for example, a combined state might be "female with a part-time job, male with a full-time job, one child". The advantage of this second alternative -the one we chose -is to take into account the interdependency of the dimensions right from the coding stage.…”

Welfare States’ Policy Response to New Social Risk: Sequence Analysis for Welfare Policy

“…Sequence analysis, as discussed in detail in following section, is a research method designed to compare the trajectories of data and uncover its underlying patterns (Kruskal 1983;MacIndoe and Abbott 2009). Often utilized in the analyses of DNA, it was originally developed to emphasize (in contrast to variable-based causal models) a longitudinal, holistic approach, and the development of explanatory typologies (Pollock 2007). In a manner complementary to studies of institutional change, sequence analysis can effectively discern whether welfare states maintain existing institutions despite external change (drift), attach new elements to existing institutions (layering), redeploy the old system to a new purpose (conversion) (Streeck and Thelen 2005), or functionally recalibrate existing institutions (Hemerijck 2013).…”

“…To calculate the substitution costs, a substitution matrix is used based on transition likelihoods between positions 7 (Caswell and Kleif 2013;Han and Moen 1999). Transition likelihoods represent a data-generated definition of costs in opposition to a theoretically derived definition and are recommended when lacking specific theory on the subject matter (Lesnard 2006;Pollock 2007; see also Caswell and Kleif 2013). This paper also uses a transition likelihood substitution matrix because it is difficult to explain theoretically the difference of 6 The quartiles are designed on the basis of the data of 21 countries and 30 years (630 data points) by each variable.…”