Mapping Isolation Regimes: A Latent Class Analysis of Institutional Sorting in Solitary Confinement

• In Event: Unpacking Solitary Confinement: Isolation Types, Institutional Sorting, and Reform

Abstract

Research on solitary confinement typically relies on binary indicators that obscure variation in how isolation is organized and experienced within prisons. We reconceptualize solitary confinement as a heterogeneous set of institutional placements and examine how prisons sort individuals into distinct isolation regimes. Using nationwide administrative data from Danish prisons (2010–2023) linked to longitudinal registry data, we apply latent class analysis (LCA) to identify empirically distinct isolation types and then model predictors of class membership. Extending this approach, we identify higher-order classes capturing patterned combinations of isolation experiences over the course of incarceration and estimate risk factors associated with these trajectories. The results reveal multiple, systematically structured isolation regimes and show that individual characteristics, prior criminal justice contact, and institutional pathways differentially predict both single-type and multi-type confinement patterns. By moving beyond dichotomous measures, this study advances criminological research in two ways: it empirically organizes solitary confinement into distinct institutional types and demonstrates how penal institutions stratify confinement through patterned sorting mechanisms, contributing to broader debates on classification, governance, and penal stratification.

Authors

• Christopher Wildeman, Duke University

• Frederik Rom Taxhjelm, University of Copenhagen

• Stinne Skriver Jørgensen, The Danish Institute for Human Rights

• Lars Højsgaard Andersen, ROCKWOOL Foundation