The Republican National Committee filed yesterday an emergency stay application in the PA Genser litigation. The theory underlying the RNC brief significantly expands the scope of absentee ballots potentially affected. In light of that, I have to update my initial estimate, with significantly more ballots potentially at stake.
The Genser litigation involves the specific issue of absentee ballots that might be defective because voters failed to place their ballots in the required secrecy sleeve. The issue is whether those voters can cast a provisional ballot on election day, which is one means to ensure they do not lose their right to vote altogether. But the RNC’s position before the Supreme Court is that state law prohibits voters whose absentee ballot is defective for any reason from casting a provisional ballot. This would include ballots that lack a proper date or signature on the ballot envelope.
My initial analysis was based only on the number of potentially defective ballots due to lack of a secrecy sleeve. As the PA supreme court said, “the issue before us is narrow; it asks the Court to consider the effect of a naked mail-in ballot…” But now that the RNC has made clear that it seeks to put all defective absentee ballots at stake, those numbers change. Based again on this important academic study by Marc Meredith, Michael Morse, Amaya Madarang, and Katie Steele of the 2022 PA elections, it appears that 1.4% of the absentee ballots returned in 2022 were rejected as defective for any of these reasons (I’m leaving aside ballots rejected for arriving too late). Note also there’s reason to think there might be a bit of under-reporting in the PA data base.
The PA counties vary in whether they provide notice to a voter that their absentee ballot is being rejected and an opportunity to fix that defect. Depending on the county, there are various ways to do that. The study examines whether people whose mail ballot is rejected go on to cast a vote successfully, either by “by mail” or “in person.” Of all the voters with rejected absentees due to the secrecy sleeve issue or problems with their ballot envelope, 13.7% ended up voting in person.
I’m assuming that roughly 1.8 million voters will return an absentee ballot this year (I’ll leave aside the basis for that, but it’s based on current absentee request rates). That would mean around 25,200 rejected absentees. If the same percentage of those voters would vote in person, through a provisional ballot or otherwise, that would be 3,452 votes. Remember that these voters will not all vote for one candidate. If we assume that Harris wins 2/3rd of these voters, that means the new bounds we can put on the potential partisan effect look like this: if everyone who returned an absentee ballot that was rejected for ballot envelope defects or lack of secrecy sleeve would vote use a provisional ballot in person if permitted, this would mean that Harris would win a net of 8,316 votes from these ballots. On the other hand, if the same percent of these voters would vote in person in 2024 as did in 2022, Harris would net 1,139 votes from these voters.
That gives an updated range of the potential number of votes at stake. Again, there are a lot of assumptions built in here, and I’m prepared to revise further if more detailed information comes to light.
I should note that the RNC brief relies on one source of data from the 2020 election, which estimates that 1% of absentee ballots were rejected for lack of a secrecy sleeve (that source notes possible incompleteness in the data). The RNC, of course, has a stake in presenting a maximal view of how many ballots are potentially at stake, in order to convince the Court of the importance of the issue. That’s a significantly higher percentage than in 2022 and there are strong reasons to believe the 2022 figures are more reliable. In 2019, PA adopted no-excuse absentee voting for the first time. Most voters who voted absentee in 2020 were doing so for the first time – for that reason and because the pandemic led to a massive expansion in absentee voting. In addition, it was not until Sept. 2020 that the PA supreme court ruled that ballots without a secrecy sleeve were invalid. The campaigns and state officials made major efforts to explain this to voters, but with little time to educate them.
We know that first-time absentee voters make a mistake on their ballots at much higher rates than prior absentee voters. Many voters who vote absentee this year will have voted absentee at least once before. To be sure, there will also be voters doing so for the first time, but that number will be significantly less than in 2020 (on the other hand, mid-term voters are perhaps less likely to make mistakes than general election voters, so that factor could somewhat offset these other factors).
Now that the more expansive RNC position in the PA litigation became clear, I’ve attached an update note to my prior post.