Abstract The goal of this paper is to study the electoral impact of crisis management policies. With this aim, we exploit a natural experiment during the COVID-19 pandemic in France to evaluate the effect of the lockdown on voting behaviour. In particular, the country has been divided in two areas, red and green, subject to a ‘hard’ and a ‘soft’ lockdown, respectively. To measure voting behaviour, before and after the policy, we rely on the 2020 French municipal elections: the first round took place before the introduction of the restrictions, while the second round was delayed after the end of the lockdown. We estimate a spatial regression-discontinuity-design model comparing electoral outcomes around the border of red and green areas both in the second round and between the two electoral rounds. The main results suggest that lockdown regulations significantly affected voting outcomes. First, in localities under a harder lockdown, the incumbent’s vote share is higher. Second, voter turnout is larger where more stringent restrictions are adopted. These results suggest that lockdown policy mobilizes citizens and leads them to rally around the incumbent politicians.