Pull-forward and induced vaccination under time-limited mandates: Evidence from a low-coercion mandate

Vaccine mandates featuring a deadline, i.e. time-limited, can raise uptake either by pulling forward vaccinations that would have occurred later or by inducing additional vaccinations that would not have occurred absent the mandate. This paper asks how such mandates change vaccination behaviour, how the overall effect decomposes into the pull-forward and induction components, and which features of the mandate and public-health context drive that composition. Empirically, we study a low-coercion time-limited mandate targeting graduating high school students in Western Australia and identify its causal effects using regression discontinuity designs based on strict school-age eligibility rules, applied to population-wide administrative records on first-dose COVID-19 vaccinations (linked AIR-PLIDA data, 1 June 2021-21 May 2022; analysis sample of 46,421 Year 11-12 students). We estimate both a static RDD at the deadline and a dynamic RDD that estimates the treatment effect over time. The mandate increased short-run first-dose uptake by 9.3 percentage points (12.7%) among the targeted cohort, but the dynamic evidence shows that this effect is entirely driven by pull-forward behaviour: uptake converges in the long run, implying no vaccinations were induced. Students advanced vaccination by up to 80 days. Theoretically, we develop a simple present-bias model of vaccination under deadlines. We use it to interpret the empirical patterns and to derive, among other results, conditions under which time-limited mandates are more likely to pull forward vaccinations rather than induce them. Our findings highlight the importance of evaluating mandates beyond short-run windows and provide a framework for designing and interpreting time-limited vaccination policies.