Hongli Zhao
Whether digital infrastructure development improves population health outcomes remains empirically contested. This study exploits China's "Broadband China" pilot policy-implemented in staggered batches across 117 cities during 2014-2016-as a quasi-natural experiment and employs a difference-in-differences (DID) design with seven waves of the China Family Panel Studies (CFPS, 2010-2022) to estimate the causal effect of broadband infrastructure expansion on individual body weight and weight-related health inequality. The main findings are threefold. First, the "Broadband China" policy significantly reduces individual BMI by 0.254 units and lowers the BMI relative deprivation index by 0.004, indicating simultaneous improvements in average body weight and reductions in health inequality. Second, heterogeneity analyses reveal a dual-edged pattern: the policy disproportionately benefits vulnerable populations-including residents in economically underdeveloped regions, women, and older adults-thereby narrowing digital health divides, yet exerts stronger effects among highly educated groups, suggesting that educational disparities may perpetuate inequality through differential digital health literacy. Third, mechanism analysis reveals that the policy functions through four pathways: improving mental health, enhancing life satisfaction, cutting working hours and raising physical activity time. In terms of negative competitive mechanisms, the policy also exerts adverse effects by delaying sleep time and increasing sedentary duration. These findings provide policy-relevant evidence for leveraging digital infrastructure as a public health intervention tool in developing countries.