Do State Insulin Price Caps Improve Diabetes Management? Evidence from Staggered Policy Adoption

Abstract

Between 2020 and 2023, eighteen U.S. states enacted laws capping monthly out-of-pocket costs for insulin, typically at \$25–\$100. While prior research documents reduced consumer spending under these caps, no study has examined downstream health outcomes. This paper provides the first quasi-experimental evaluation of state insulin price caps on diabetes management using Behavioral Risk Factor Surveillance System (BRFSS) data from 2019–2022 and a staggered difference-in-differences design. I find a small, imprecise overall treatment effect on insulin use rates: the Callaway-Sant'Anna ATT is 1.9 percentage points (SE = 1.3 pp, 95% CI: $[-0.7, 4.5]$). Event study estimates show point estimates growing over time (5.1 pp at $t+2$), but also reveal pre-treatment coefficients that raise concerns about the parallel trends assumption. A Medicare placebo test supports the identification strategy—state caps show no effect on Medicare beneficiaries (who are unaffected by state-regulated plan rules), while under-65 populations show marginally significant effects (ATT = 3.0 pp, 95% CI: $[0.03, 6.0]$). However, Rambachan-Roth sensitivity analysis demonstrates that even under perfect parallel trends, the main effect is too imprecise to reject a null. These findings suggest either no effect or a modest positive effect of insulin price caps on insulin uptake among diabetics, with important limitations on causal interpretation. The results highlight the need for longer post-treatment periods and individual-level administrative data to credibly evaluate this rapidly expanding policy.