Mapping son preference in India, 2002–2021: Spatial patterns and trends using model-based small area estimation

Background: Despite widespread interest in son preference in India, the study of its spatial distribution and trends by parity at the district level is limited. Objective: This study investigates spatial patterns and temporal trends in son preference by parity across districts of India from 2002 to 2021. Methods: We applied model-based area-level small area estimation techniques on data from consecutive rounds of the Indian National Family Health Survey and the District Level Household Survey to derive district-level estimates of son preference by parity. Spatial patterns and clustering were examined using Moran’s I and local indicators of spatial autocorrelation across multiple rounds of survey. Results: At parity 1, only a few districts showed a strong son preference in 2002–2004, but this number steadily increased in each subsequent survey round. In contrast, the number of districts with high son preference at parity 2 rose in 2007–2008 and then declined in subsequent rounds. At parity 3 or higher, the number of districts showing a strong son preference declined consistently in each subsequent survey round. Conclusions: Son preference in India exhibits distinct spatial and evolving temporal patterns across parities. The increasing prevalence at parity 1 and declining trends at higher parities suggest shifting reproductive behaviours. These findings underscore the need for targeted district-level, parity-specific interventions to address persistent and emerging gender-biased norms in son preference. Contribution: This study is the first of its kind to examine trends in son preference by parity across Indian districts in the last two decades using data from four consecutive rounds of large-scale national household surveys conducted between 2002 and 2021.