This preprint presents an empirical assessment of a theorem-driven framework that conceptualizes crime as a mediated, gated, and path-dependent process. Building on a separate theoretical paper, the study examines whether recurring empirical patterns in criminological and policy-relevant research are more coherently interpretable when crime is understood as conditioned by social mediation, governance constraints, and temporal persistence rather than as a direct outcome of criminogenic exposure. The article employs a theory-driven qualitative secondary analysis of a purposive corpus of empirical studies addressing individual offending processes, gang entry and exit, community social order, institutional cooperation, crime prevention, governance capacity, and long-term intervention outcomes. The aim is not strict hypothesis testing or direct verification, but analytical assessment: whether heterogeneous empirical findings become more intelligible when read through the three core theorems of the broader framework. The study is organized around three propositions:(1) criminogenic exposure is not analytically equivalent to realized crime,(2) social cohesion and preventive capacity are not freely deployable but gated by governance, trust, and feasibility conditions, and(3) crime-related patterns may persist over time through path-dependent mechanisms even where upstream conditions improve. Across the empirical material, the findings are broadly consistent with this framework. The analysis indicates that crime is more adequately understood as a conditional and temporally structured process shaped by mediation, rationalization, institutional constraints, and persistence mechanisms, rather than as a straightforward behavioral output of structural disadvantage or exposure alone. The article contributes theoretically by linking abstract structural propositions to observable empirical indicators across multiple levels of analysis, and methodologically by demonstrating how theorem-driven qualitative secondary analysis can be used to assess the empirical coherence of a general theoretical model. This work is released as a preprint and is distributed under the Creative Commons Attribution 4.0 International (CC BY 4.0) license.
J. E. Fröderberg (Fri,) studied this question.