The effect of lockdowns on mental health: evidence from a natural experiment analysing an Australian longitudinal probability sample survey

Abstract

Background. Many studies have examined population mental health during the COVID-19 pandemic but have not been able to isolate the direct effect of lockdowns. Australia provides a rare opportunity to examine lockdown effects because its aggressive suppression policy response to COVID-19 enables disentangling the introduction of lockdowns from high rates of community transmission. Methods. Difference-in-difference models were estimated that compared changes in mental health of those exposed to lockdown with those not exposed. Mental health was assessed using the Mental Health Inventory (MHI-5). The analysis was possible because of the coincidence of annual national longitudinal survey data collection during 2020 with a natural experiment, with residents in one part of Australia (Victoria) exposed to lockdown and those living elsewhere relatively free of restrictions. Outcomes. A significant, but relatively small, treatment effect of the lockdown on MHI-5 was found (-1·4 points: 95% CI -1·7 to 1·2). The treatment effect was larger for females ( 2·2: 2·6 to 1·7) and even larger for coupled females with dependent children ( 4·4: -5·0 to -3·8) and for females without access to personal outdoor space at home ( 4·2: 5·5 to 2·9). Treatments effects were generally small for males ( 0·7: 0·8 to 0·5). Interpretation. The imposition of lockdowns had a modest negative effect on overall population mental health. The mental health burden of lockdowns was not, however, felt equally by all groups.