June 2021

IZA DP No. 14481: The Effects of COVID-19 on Employment, Labour Markets and Gender Equality in Central America

This study considers the economic impact of Covid-19 on enterprises in four Central American countries – El Salvador, Guatemala, Honduras and Nicaragua. At the time of the analysis neither the pandemic nor its economic consequences had fully run their course. It is not, therefore, a definitive analysis but it is important to try to draw important lessons as soon as possible. The main focus of the study was the initial impact on labour markets. The analysis was based on World Bank enterprise surveys undertaken before the outbreak of Covid-19 and follow-up surveys on the effects of the pandemic, also undertaken by the World Bank. These were combined with data on government containment measures and on morbidity and mortality rates. The use of enterprise data to analyse labour market issues has some limitations but also many strengths. The data is useful for analysing the consequences for gender equality in employment. Since the demand for labour is a derived demand firm level data provides a clear link to labour market effects. The pandemic has caused a significant loss in sales for many firms, This creates a loss of liquidity which, in turn, has caused some firms to reduce employment, working hours and wages. Government containment measures necessary to save lives such as temporary workplace closures have added to the burden for both firms and employees. The study starts by using the surveys to identify the important stylised facts. Although some issues are already well documented anecdotally through media reports this provides a more evidence based approach. It also helps identify several issues, such as the impact on gender equality which have received less journalistic attention. The study is further supported by a regression analysis (OLS and SURE) of several key outcomes (changes in sales, employment, the share of females in employment and firm expectations of survival). A limitation of such analysis with any enterprise level is heterogeneity and, in consequence, a risk of sample selection bias. To provide robustness checks we use a matching approach. The results suggest that a significant proportion of surviving firms are vulnerable to permanent closure. The ability of firms to retain labour depends on sales which are affected by both the pandemic itself and the government containment measures. Only a small proportion of firms have received government support and there is evidence that it could help both firm survival and the retention of labour. There is some doubt whether the four countries have the institutional capacity to provide effective support. If such doubts prove well founded then support may need to be externally driven.