Recent global adversities have illuminated the shortcomings of a singular mortality benchmark for identifying famine—a methodology that risks obscuring the actual progression of starvation across diverse populations. In a scholarly publication featured in the Lancet, researchers affiliated with Columbia University Mailman School of Public Health, alongside their collaborators, advocate for a profound re-evaluation of the established criteria for famine classification.
The established mortality benchmarks employed by the Integrated Food Security Phase Classification (IPC) were initially devised for rural African contexts, rather than for urban demographics in middle-income countries. Significant discrepancies are evident in the assessment of famine-related mortality depending on the specific setting.
L.H. Lumey, MD, PhD, Columbia Mailman School professor of Epidemiology
Dr. Lumey and fellow authors contend that widespread starvation can persist without official designation for extended durations if it fails to meet the IPC’s Phase 5 threshold of two fatalities per 10,000 individuals daily. Consequently, mass starvation events may remain undetected until they have reached an advanced stage.
Furthermore, the researchers highlight that the IPC primarily considers absolute mortality rates, neglecting significant proportional increases within particular age demographics. Empirical data derived from the Dutch Hunger Winter—a period extensively researched by Dr. Lumey—serves to underscore this proposition. This historical event was characterized by a decline in birth weights and a reduction in birth rates, subsequently followed by a substantial surge in child mortality. Within major urban centers, infant mortality in March 1945 escalated to four times its pre-war levels, while mortality among children aged one to four years experienced a sevenfold increase. Nevertheless, as Dr. Lumey noted, “these marked increases would not qualify under the prevailing IPC famine criteria for children under the age of five.”
Moreover, mortality serves as an inherent retrospective indicator. By the time these established benchmarks are met, preventable deaths due to starvation have already transpired. The process of classification can also be susceptible to political influence, given that access to accurate mortality data is frequently impeded or subject to distortion.
“The identification of more sensitive early indicators of famine stress could expedite the interval between acute food insecurity and escalating mortality rates,” stated Dr. Lumey. “Adopting a more nuanced and contextually relevant approach would facilitate more prompt humanitarian interventions.”
Additional contributors to this research include Ingrid de Zwarte, affiliated with Wageningen University; and Alex de Waal, associated with Tufts University.
de Zwarte, I., et al. (2026). Rethinking current famine classification: insights from history. The Lancet. DOI: 10.1016/s0140-6736(26)00214-x. https://www.thelancet.com/journals/lancet/article/PIIS0140-6736(26)00214-X/fulltext
