COVID-19 and the labour market: - Estimating the employment effects of South Africa’s national lockdown

We then use the timing of the national lockdown, and the timing of the QLFS data 4 COVID-19 and the labour market: Estimating the employment effects of South Africa’s national lockdown collection interviews, to estimate difference-in-differences (DiD) models on a matched panel sample. [...] ? ? 3 is the main coefficient of interest, as it measures the causal effect of the onset of lockdown policy; that is, the average difference in outcomes between the treatment and control groups in the post-treatment period relative to the pre-treatment period. [...] Notes: [1] Figure presents OLS coefficients of interaction terms of treatment and period from the DiD regression of the effect of the lockdown on the probability of employment, without controlling for individual FEs or the vector of control variables. [...] In this paper, in addition to providing a detailed descriptive account of the effects of the COVID-19 pandemic on employment in the South African labour market, we exploit quasi-experimental variation in the country’s national lockdown policy to estimate the causal effect of the country’s lockdown policy on the probability of employment. [...] 22 We arrive at this estimate by augmenting the percentage reduction in employment exhibited among the control group (5%: what the treatment group would have experienced in the absence of treatment under the parallel trends assumption) with the estimated ATT of 8 percentage points, and then using this augmented estimate (13%) we calculate the reduction in employment among the treatment group and m.