NEW: My @MPSAnet working paper with @MindySRomero and Jennifer Paluch looks at the effect of vote-by-mail (VBM) access policy on turnout and presidential vote in 2020. A thread with our findings. 1/
President Trump argued easing VBM would increase turnout at the expense of Republican candidates. Existing research suggests some VBM policies do increase turnout but do not in fact help or hurt either party. 2/
These conclusions might be different for 2020 amidst the threat of COVID, the messiness of the pandemic election, and the politicization of mail balloting. So we use the same measurement strategies others have employed in pre-pandemic studies to see if things changed in 2020. 3/
FOR TURNOUT: We found strong positive effects of mailing every voter a ballot, esp in a hypothetical zero-VBM county, and that didn’t really change in 2020. Sending VBM applications had a smaller positive effect, also both before and during 2020. 4/
Removing excuses for VBM had a small positive turnout effect before 2020, but it had a negative effect in 2020 (more on that in a sec) 5/
FOR PRESIDENTIAL VOTE: We found no partisan effect of mailing ballots, but contrary to other research we found weak pro-R effects before 2020. We also found even larger pro-R effects for all 3 policies in 2020. So more of our models showed a pro-Trump effect than anything. 6/
IMPORTANT CAVEAT: We controlled for each county’s trend over time because we found evidence of such trends, which would violate core assumptions of our modeling approach unless they’re accounted for. So all these effects are measured against these longer-term trends. 7/
The turnout effect of mailing ballots is robust to that decision, but the rest of our findings are much more sensitive to it, and to some other modeling decisions as well. 8/
Our conclusions? Mailing every voter a ballot increases turnout; the other reforms have far smaller effects. The partisan effects are challenging to estimate but overall point to no effect or maybe even a small effect that favors Republicans. 9/