Responding to the call: Population affinity estimation in a South African forensic context of identification - are we helping or harming?

Abstract

There is growing momentum in forensic anthropology to move away from using socially ascribed racial groups in analysis and reporting. This study evaluated whether and how anthropological estimation of population-affinity influences identification outcomes in forensic cases processed at the Forensic Anthropology Cape Town (FACT) laboratory in South Africa. A total of 172 cases (174 individuals) were examined. Estimated South African population-affinity distributions were: 37 % Mixed, 22 % African, 4 % European, 12 % displaying traits from multiple affinities, and 25 % indeterminate. Identification outcomes were assessed for 168 cases (six were known prior to assessment). Overall, 37 % were positively identified, 49 % remained unidentified, and 14 % had unknown status. Identification rates varied: African (51 %) and European (83 %) population-affinity groups showed higher rates of positive identification compared to Mixed (26 %) and multiple affinities (20 %). Notably, individuals with indeterminate affinity achieved a 44 % identification rate. Statistical testing indicated no significant association between estimated population-affinity and identification outcome (p = 0.974). Where comparison was possible, estimated population-affinity aligned with socially ascribed racial categories in 84 % of cases, with six misclassifications. Challenges were noted in distinguishing between South African Mixed and African population-affinities. Although the sample size was limited, our findings support discontinuing the inclusion of population-affinity estimation in South African forensic anthropology reports. In practice, law enforcement often translates anthropological affinity estimates into social racial categories, which in the South African context risks misclassification and undermines identification efforts.