Authors: Giorgio Cutuli and Raffele Grotti
In the past years, an increasing amount of literature has shown a set of spillover and feedback effects between poverty, low pay and unemployment (Gallie et al. 2003; Stewart, 2007; Andriopoulou and Tsakloglou, 2011; McKnight et al. 2016 ). Moreover, even if to a different extent depending on the specific social group and institutional setting, low-pay, poverty and unemployment spells tend to persist and/or to reiterate over time (Biewen, 2014; Gangl, 2006).
The micro-level stickiness of these disadvantaged conditions has a strong impact in terms of social stratification, as the severity of socio-economic risks for individuals and households is positively associated with the duration of such negative experiences (Janlert et al. 2015; Schmieder et al. 2016). Therefore, it can be argued that analysing the longitudinal character of socio-economic risks can provide important insights.
Building on the methodological literature on ‘dynamic correlated random effects models’ and ‘initial condition problem’ (Wooldridge 2005; Hesketh and Skrondal; 2013), a new paper by Giorgio Cutuli and Raffaele Grotti proposes an innovative framework to be used in the analysis of drivers of the accumulation of several socio-economic risks. Namely, their dynamic approach allows analysing alternative patterns of stratification and accumulation of socio-economic disadvantages over time when the modelled outcome is dichotomous.
Specifically, the authors use longitudinal data from the European Union Statistics on Income and Living Conditions (EU-SILC) for Denmark, France, Italy and the United Kingdom for the years 2004 to 2015. Through dynamic models, they disentangle ‘genuine state dependence’ (GSD) and ‘unobserved heterogeneity’ (UH) in individual characteristics as possibly distinct, yet mutually reinforcing, (i.e. interactive) drivers.
GSD refers to the net scarring effect of previous unemployment (t-1) on current unemployment (t), thus leading to dynamics of accumulation over time. UH refers to a set of unobserved individual and time-constant characteristics, including ability and motivation, which are positively correlated with the outcome.
These unobserved factors are associated with unemployment exposure at any t, irrespective of the previous unemployment status, so that individuals with the highest levels of UH positively correlated with unemployment (e.g. with the lowest ability and the lowest motivation) experience higher unemployment risks consistently over time.
The authors investigate heterogeneity in unemployment dynamics and show their variation across workforce segments. Their analysis indicates that both unobserved heterogeneity and genuine state dependence are relevant in accounting for the stratification and the accumulation of unemployment and long-term unemployment risks over time, both in additive and in multiplicative terms. Besides, the analysis leads to a twofold conclusion.
First, the (relative) weight of UH and GSD, as well as their interplay vary across different labour market and institutional settings (i.e. countries), thus indirectly confirming how institutional and contextual heterogeneity can shape the magnitude and dynamics of the accumulation of occupational risks.
Second, GSD varies significantly across units having distinct unobserved characteristics correlated with unemployment, leading the weakest segment of the workforce (i.e. those having a past history of unemployment) to unevenly shoulder the burden of multiplicative dynamics of cumulative disadvantage.
The distinction between GSD and UH, and the conceptualization of their interplay, comes with theoretical implications. The study provides useful insights concerning a more fine-grained understanding of cumulative disadvantaged dynamics and enriches existing perspectives in the study of longitudinal processes of social stratification.
In light of this evidence, the contribution suggests how the concept of ‘scarring effects’, especially when used to identify the recurrence and life-course accumulation of disadvantages, can be conveniently supplemented with the concepts of unobserved heterogeneity and genuine state dependence as interactive drivers of protracted socio-economic risks and social stratification.
Moreover, their findings suggest that labour market and welfare policies would be more effective if designed and targeted to specific social groups according to the relative importance of genuine state dependence in shaping unemployment, or more generally, socio-economic risks inertia.