Ten years on: are raw material criticality assessments making more sense?

Raw material criticality assessments are essential tools for policy-makers, industry stakeholders, and researchers. A decade ago, two articles critiqued the compliance of criticality studies with risk theory, concluding that raw material criticality assessments were fundamentally flawed in several ways. Here, we systematically review 36 peer-reviewed criticality assessments published between 2017 and 2024, examining whether they have corrected previously identified methodological short comings. We show that most studies rely on unjustified aggregation methods, some of which may yield statistically “worse-than-useless” results, and consistently overlook key aspects of quantitative risk theory, such as the spectrum of severity and duration of the disruption. Empirical validation of indicators is rare, revealing continued reliance on indicators without demonstrated causal links. We also find that nearly half of the studies citing the two articles criticizing the field from the point of view of risk theory misinterpret or overlook the core arguments of these critiques. We complement this review by identifying methodological advancements from adjacent disciplines, including quanti tative risk assessments of natural hazards, empirically grounded causal modeling, and input-output analyses. Drawing on these advancements, we recommend that future work (i) specify scope unam biguously—whose risk, over what horizon, and for which disruption mechanisms; (ii) express risk in physical or economic units; (iii) calibrate models against historical disruption data; (iv) employ proxies only when their causal link to the target quantity is demonstrated; and (v) adhere to mathe matically correct, dimensionally consistent formulations. Implementing these principles would yield decision-relevant, empirically robust insights for industry and policy.