Universal vote-by-mail has no impact on partisan turnout or vote share

Source: Daniel M. Thompson, Jennifer A. Wu, Jesse Yoder, and Andrew B. Hall, PNAS, first published June 9, 2020
From the abstract:
In response to COVID-19, many scholars and policy makers are urging the United States to expand voting-by-mail programs to safeguard the electoral process, but there are concerns that such a policy could favor one party over the other. We estimate the effects of universal vote-by-mail, a policy under which every voter is mailed a ballot in advance of the election, on partisan election outcomes. We find that universal vote-by-mail does not affect either party’s share of turnout or either party’s vote share. These conclusions support the conventional wisdom of election administration experts and contradict many popular claims in the media. Our results imply that the partisan outcomes of vote-by-mail elections closely resemble in-person elections, at least in normal times.