Social Multipliers in Vaccination

Social multipliers resulting from peer effects are important for policy analysis because they imply that aggregate effects of interventions are larger than individual effects. Social multipliers for vaccination may arise due to information transmission among providers and patients, diffusion of the vaccine among providers, social norms to vaccinate, and herd immunity. This paper tests for the presence of social multipliers for HPV, influenza, and TDAP vaccination. We analyzed Behavioral Risk Factor Surveillance System (BRFSS) data for adults aged 18-49 in counties with population greater than 10,000 in states that used the HPV module (9 states) and National Immunization Survey-Teen (NIS-Teen) data for teens aged 13-17 from all 50 states and the District of Columbia for the years 2008-2014. The first estimate of the social multiplier (“excess variation”) used the ratio of the observed area-level standard deviation of vaccinations to the standard deviation predicted under the assumption of independent and identically distributed vaccinations. Social interactions imply that the variance across groups is greater than the expected variance without social interactions (i.e., the ratio, or social multiplier, is greater than 1). The second estimate of the social multiplier used a regression of areal-level mean outcomes on the mean of predicted outcomes from an individual-level regression. With social multipliers, aggregate coefficients will be bigger than individual coefficients and the estimated coefficient, which provides an estimate of the social multiplier, will be greater than 1. Both tests were conducted with and without regression adjustment for covariates measuring demographics, access to health care, area-level characteristics, and year indicators. Across both tests at the county and state level, this study finds weak to no evidence for social multipliers in vaccination. To the extent that social multipliers are present, interventions that increase HPV vaccination could have positive spillovers at the population level. However, uptake of newer vaccinations like HPV may not have reached the “tipping point” yet, especially for boys.