This chapter uses advanced meta-analysis techniques to evaluate the short-run impact of COVID-19 on various labour market indicators. Using 2429 reported estimates of labour market outcomes associated with COVID-19 from 29 empirical studies conducted in 12 countries, we show that large parts of the documented literature exhibit substantial publication bias. After controlling for publication bias, we find almost no practically meaningful impact of COVID-19 on earnings, hours worked and (un)employment. Next, we investigate if the identified publication bias is caused by publication characteristics. However, these characteristics do not appear to drive the identified bias. We also uncover the differences between developed and developing countries. The findings indicate that in practical terms both groups of countries experienced hardly any measurable short-term impacts of COVID-19 on the labour market. Furthermore, we detect systematic upward publication bias for unemployment and job loss in developed countries, and downward bias for employment, hours worked and earnings in developing countries. Overall, we concluded that in the short-run no major labour market effects could be identified related to formal employment.
|Title of host publication||Covid-19 and International Development|
|Pages||71 - 88|
|Publication status||Published - 2022|