ABSTRACT
This paper elucidates the causal effect of lockdowns on social distancing behaviour in Turkey by adopting an augmented synthetic control and a factor-augmented model approach for imputing counterfactuals. By constructing a synthetic control group that reproduces pre-lockdown trajectory of mobility of the treated provinces and that accommodates staggered adoption, the difference between the counterfactual and actual mobility of treated provinces is assessed in the post-lockdown period. The analysis shows that in the short run following the onset of lockdowns, outdoor mobility would have been about 17-53 percentage points higher on average in the absence of lockdowns, depending on social distancing measure. However, residential mobility would have been about 12 percentage points lower in the absence of lockdowns. The findings are corroborated using interactive fixed effects and matrix completion counterfactuals that accommodate staggered adoption and treatment reversals.
ABSTRACT
Introduction: Social or physical distancing has been an effective measure for reducing the spread of COVID-19 infections. Investigating the determinants of adherence to social distancing can inform public health strategies to improve the behaviour. However, there is a lack of data in various populations. This study investigates the degree to which South Africans complied with social distancing during the country's COVID-19 lockdown and identifies the determinants associated with being in close contact with large numbers of people. Materials and Methods: Data was collected from a South African national online survey on a data free platform, supplemented with telephone interviews. The survey was conducted from 8 to 29 April 2020. The primary outcome was the number of people that participants came into close contact with (within a 2-metre distance) the last time they were outside their home during the COVID-19 lockdown. Multivariate multinomial regression investigated the socio-demographic, psychosocial and household environmental determinants associated with being in contact with 1-10, 11-50 and more than 50 people. Results: Of the 17,563 adult participants, 20.3% reported having not left home, 50.6% were in close physical distance with 1-10 people, 21.1% with 11-50 people, and 8.0% with >50 people. Larger household size and incorrect knowledge about the importance of social distancing were associated with being in contact with >50 people. Male gender, younger age and being in the White and Coloured population groups were significantly associated with being in contact with 1-10 people but not with larger numbers of people. Employment, at least secondary school education, lack of self-efficacy in being able to protect oneself from infection, and moderate or high risk perception of becoming infected, were all associated with increased odds of close contact with 1-10, 11-50, and >50 people relative to remaining at home. Conclusion: The findings identify subgroups of individuals that are less likely to comply with social distancing regulations. Public health communication, interventions and policy can be tailored to address these determinants of social distancing.