Intervention to Encourage Voting by Mail Worked, Researchers Say
Pennsylvania voters who received a postcard encouraging them to vote by mail in the state’s June 2 primary were more likely to request a ballot and successfully cast it than voters who did not receive the reminder, a University of Pennsylvania study found.
Pennsylvania was one of several states that moved its primary due to the onset of the coronavirus pandemic last spring.
Election officials had originally planned to participate in a kind of northeast Super Tuesday, with voters heading to the polls along with their counterparts in Connecticut, Delaware, Maryland, New York and Rhode Island.
Pennsylvania then joined several other states in moving its primary to June 2.
About two weeks before the new date, University of Pennsylvania researchers were joined by Philadelphia officials in randomly sending 46,960 of 935,745 registered voters a postcard encouraging them to vote by mail.
Researchers Daniel J. Hopkins, Marc Meredith, Anjali Chainani, Nathaniel Olin, and Tiffany Tse said registered voters who received the postcard were 3% more likely than those who didn’t to successfully cast their ballot.
According to a separate study, “Reconsidering lost votes by mail,” the share of ballots cast by mail increased six-fold over prior elections in several states holding elections between April and June 2020.
“While increased access to mail ballots has helped protect voting rights, there are also concerns that their increased use could disenfranchise voters,” the authors of the new study wrote.
They noted that a mail ballot may be less likely to count than a ballot cast in person for multiple reasons:
1) Mail ballots may not be received in time;
2) mail ballots may have higher rates of clerical errors; and
3) the process of casting an in-person ballot may identify and rectify errors.
“The share of mail ballots affected by these issues likely increased during the COVID-19 pandemic. Partly, that is because people who would typically cast an in-person ballot are more likely to cast a problematic mail ballot than those experienced with voting by mail,” the authors said. “Also, election officials are more likely to face issues distributing and tabulating mail ballots when there is an upsurge in mail ballots.”
They also noted that, as this was the first statewide federal election held since Pennsylvania adopted universal access to mail ballots, information about how to apply was particularly likely to increase awareness of voting by mail.
The research built on a 2019 study of the city of Philadelphia in which voters were sent multiple postcards over several months.
That study found that sending a registered voter a prefilled postcard substantially increased the rate at which registrants requested mail ballots.
The new study codified those results, finding that 35% of the increased voting by mail came from substitution by people who would have otherwise voted in person.
While Pennsylvania state law specifies that mail ballots must be received by Election Day, issues with mail ballot distribution and large-scale protests following George Floyd’s death led Pennsylvania’s governor to order that ballots postmarked by Election Day be counted if received within a week of Election Day.
“The postcards conveyed information about the May 26 deadline to request a mail ballot and included a message either encouraging voters to request a ballot because ‘[i]t’s safer for you to vote by mail!’ or ‘[i]t’s safer for [neighborhood] to vote by mail!”
Among other things the team looked at was whether mail in balloting skewed the voting population to one race or another. Ultimately, they concluded it did not.
“We do not see substantively meaningful differences between respondents by imputed race,” the authors wrote.
“In Philadelphia’s June primary, encouraging voting by mail increased recorded votes by mail. Under certain conditions, mail ballots can certainly increase the use of the franchise. However, the procedures for counting mail ballots condition their ability to enfranchise.
“Mail ballots are at greater risk than ballots cast in person to be not counted because of clerical errors or procedural violations,” the authors said. “For example, in Pennsylvania’s Nov. 2020 election, thousands of mail ballots were not counted because they were not enclosed in secrecy envelopes, a provision not enforced in the primary. Also, it is plausible that primary voters were more knowledgeable about vote-by-mail options and that a similar information campaign in a highly salient general election could produce larger effects. In such conditions, political parties may well intervene to assist voters in navigating this process, too.”
“In the run-up to the 2020 elections, officials made substantial efforts through multiple media to educate voters about voting by mail. Our results provide an estimate of the impact of one such effort and may inform future efforts,” the authors said.