Are voters willing to punish incumbents for adverse changes in mental health induced by policy decisions? This paper examines this question in the empirical context of the controversial lockdown imposed by the central government of India in March 2020 in the midst of the first wave of COVID-19. Using data from three waves of a geographically representative mental health survey in India, conducted in the first, second, and third week of June 2020, respectively, we find, that among respondents who voted for the incumbent Bharatiya Janata Party (BJP) in the 2019 national election, the incidence of lockdown-induced mental health challenges was positively correlated with the intention to switch votes (if a hypothetical national election were to be held on the day of the survey) when the central government was primarily (Wave 1) or partially (Wave 2) responsible for pandemic management, and uncorrelated with the same when primary responsibility for pandemic management had visibly shifted to state governments (Wave 3). However, in contrast to the national level, the evidence for accurate responsibility attribution at the sub-national level is weak. In particular, our analysis reveals a “second order election” effect in BJP ruled states. Among the respondents in these states who voted for the BJP in the previous state elections, the incidence of lockdown-induced mental stress was positively correlated with the intention to switch votes (if hypothetical state elections were to be held on the day of the survey) when responsibility for pandemic management was primarily vested in the central government (Wave 1), and negatively correlated with the same when primary responsibility for pandemic management had visibly shifted to state governments (Wave 3).