Abstract/Summary

We estimate the importance of local labour structure in the spread of COVID-19 during the first year of the pandemic. We build a unique dataset across 6,791 English neighbourhoods which distinguishes between people living (residents) and people working (workers) in a neighbourhood, and differentiate between jobs that can be done from home (homeworkers), jobs that likely continued on-site (keyworkers), and non-essential on-site jobs. We find that a 10 percentage points increase in keyworker jobs among residents is associated with 3.15 more cases per 1000 (4.8% relative to the mean), while a 10 percentage points increase in homeworker jobs among residents is associated with a decrease of 7.74 cases per 1000 (11.8% relative to the mean). Results for the composition of workers show the same sign, but smaller magnitudes. A dynamic analysis of the monthly incidence of reported cases shows that these relationships are particularly strong during lockdown periods. These results are heterogeneous across neighbourhoods, with larger positive effect of keyworkers, and lower protective effect of homeworkers, in higher deprivation areas. We explore the role of occupation skill intensity in driving these neighbourhood differences. These findings highlight important asymmetries in the distributional impact of the policy response to COVID-19.