The population it causally identifies are people whose remote work status was affected COVID. And most of their effects are driven by those workers who became remote during COVID and live alone. People who had spouse and kids did not see their mental health worsen.