Examining the Kid Link Method as an Approach to Improve Infant Mortality Data: A Case Study in Multiracial Classification

Abstract To evaluate a novel application of the Kid Link Method (KLM) to improve racial/ethnic infant mortality statistics by addressing two common sources of error. The KLM, originally developed by the U.S. Census Bureau, assigns children’s racial classifications using parental racial information and derived child-race proportions. Multiracial infants classified as “More than one race” were examined because this growing population experiences frequent racial misclassification and data suppression. We applied the KLM to aggregated 2018–2021 Linked Birth / Death Records data across all 50 U.S. states to re-estimate deaths before age one among multiracial infants. The adjusted estimates of death counts and infant mortality rates were compared against two standard, publicly available vital data sources. The KLM identified at least 30% more multiracial infant deaths overall and expanded the number of states with unsuppressed data. Adjustment using the KLM produced an average multiracial infant mortality rate of 6.1 deaths per 1000 births, a value between the comparison sources, and increased rates by over 50% in 19 states. Discrepancies were largest in Southern states, such as Louisiana and Texas, while Western states like California and Hawaii had similar values after adjustment. Our approach uncovered the potential for significant undercounting of deaths among multiracial infants. The KLM showed potential as a method to improve data validity and coverage that could be applied to other race categories. Integrating the KLM into routine vital statistics could strengthen demographic monitoring, enhance public health data quality, and support more equitable and targeted policy responses.