Clinical trials focus in-sample (direct) effect of vaccines on health outcomes (e.g., decrease in illness, hospitalization, emergency department use, and death). Typically, however, clinical trials do not test whether the vaccines reduce transmission. FDA do not require this as it would be logistically challenging and costly to estimate indirect effects of vaccination within a clinical trial.
A new AEJ Applied Economics paper by Freedman, Sacks, Simon and Wing (2026) aims to estimate both direct and indirect effects. The use a natural experiment framework for differential vaccine eligibility roll-out by age:
Our focal state, Indiana, granted eligibility to health-care workers, first responders, and nursing home residents on December 14, 2020. In contrast to most states, however, subsequent eligibility groups in Indiana were determined only by age…Most age groups became eligible in short succession, with delays of 22 days or less between age groups, until 12–16-year-olds became eligible on May 12, 2021. Children 5–11years old did not become eligible until November 3, 2021.
The authors then examine the impact of vaccinating 11 vs. 12 year olds given the delayed vaccine access. They use two data sources: (i) Indiana Network for Patient Care (INPC) database which contains medical records from health-care providers throughout the state, and (ii) a registry of nearly all COVID lab (PCR) tests and COVID vaccinations conducted in Indiana from the Regenstrief Institute. An individual’s household location is taken from the INPC database, which includes all medical encounters (not just COVID-related encounters). The addresses are also used to input school assignment.
The direct effects of the vaccine in essence use age 11 vs. 12 status as a difference in difference approach to examine the direct impact of early access vaccination. Indirect effects are measured by looking at health outcomes for individuals aged 2–10 or 30 and older who live with either 11 years olds (who were not eligible for early vaccines) vs. 12 years olds (who were eligible for early vaccines).
The authors use a clever methodology to identify school-based indirect effects. The use the fact that some sixth graders attend middle school and others elementary school.
Our empirical strategy to identify the effects of school-wide vaccination rates on individual COVID risk is based on a comparison of sixth graders in middle schools and elementary schools. The idea behind this comparison is that middle school students in the fall of 2021 are mostly eligible for the COVID vaccine, as all students in seventh grade and higher are eligible. In contrast, elementary school students are almost all ineligible. Because some sixth graders go to middle school and others go to elementary school, this difference in eligibility leads to large differences in school-wide vaccination rates
Using these approaches, the authors find that:
Vaccination reduces cases by 80 percent, the direct effect. This protection spills over to close contacts, producing a household-level indirect effect about three-fourths as large as the direct effect. However, indirect effects do not extend to schoolmates. Our results highlight vaccine reach as important to consider when designing policy for infectious disease.
You can read the full paper here.
