How Many Votes Are Likely at Stake in the PA Litigation Over Absentee Ballots for Which the RNC is Seeking Supreme Court Review?

As Rick Hasen blogged yesterday, the RNC is going to seek Supreme Court action, based on the independent state legislature doctrine, of the PA Supreme Court decision in the Genser case. As his post explains, the PA court held that voters who cast an absentee ballot that is void because the voter failed to put the ballot inside the required secrecy sleeve can instead vote by casting a provisional ballot.

How much is at stake in this case, in terms of whether the outcome of the election could potentially be affected? I’ve done a rough estimation of that, which suggests that any partisan effect from the resolution of this issue is unlikely. I’ll address at the end some speculations about why the RNC might be pursuing the case nonetheless.

My analysis starts with this terrific article, from a group of Penn faculty, who report that in the 2022 midterms, there were 8, 072 ballots rejected for lack of a secrecy sleeve. Of those, 16% went on to vote through another means, including 10% who went on to vote in person.

In the 2020 presidential election, turnout in PA was about 30% higher than in the midterms. This was the highest turnout presidential election in modern PA history. If we assume turnout in 2024 will be just as high as in 2020 and use that same 30% increase in turnout from 2022 then, that would provide an estimate of 10,493 ballot we can estimate in 2024 that lack a secrecy sleeve (education efforts might be more effective in a presidential election, so this number might be lower; but primary voters might be more knowledgeable than general election voters, so this number might be higher).

If 10% of those voters would again seek to vote in person, that would mean we are talking about 1,049 ballots. But even then, not all of these votes would go to one candidate, of course. In 2020, when Democrats favored the use of absentees much more than Republicans, about 2/3rd of absentee votes when to Biden. I would expect that partisan gap to be narrower this time, with more Rs voting absentee this time around. But let’s assume the gap remains the same. That would mean a net for Harris of 346 votes from these voters.

Perhaps a much higher percentage of voters will vote absentee in 2024, though, than in 2022? That doesn’t seem all that likely. In 2022, 35% voted absentee. It is unlikely the percent will be significantly higher in 2024. That compares to 45% during the pandemic election in 2020. So let’s assume 40% vote absentee this time around. That would increase the net Harris votes to 394.

But let’s suppose everyone who casts an invalid absentee due to the secrecy sleeve failure would vote in person if they could. This is unlikely, based on past experience. But this would provide the maximum estimated effect of the issue. Based on the same assumptions above, this would result in a net of 3,462 votes for Harris (if 40% vote absentee, this becomes 3,956). And this is based on assumptions that maximize the number, such as the assumption that 2/3rd of absentee ballots will again favor the Ds. So based on these assumptions, that’s how close the outcome would have to be for this issue to affect the outcome — based on 2022 data on actual voting behavior, within around 400 votes, or, based on assumptions that maximize this number, around 3,956 net Harris votes.

If the margin is that size, why is the RNC so aggressively pursuing the issue? I can think of three reasons. One, that the RNC has much better information than I do and the likely effect is significantly greater than what my crude estimates and assumptions suggest. Two, in an election that might be extraordinary close, it is worth fighting over every single inch of terrain. Three, the RNC might want to try to get the Supreme Court to put some teeth into the independent state legislature doctrine and the RNC sees this as a vehicle for doing that, even if the actual number of votes in PA likely to be affected would be small.

If there are any errors in the assumptions and estimates I’ve made here, I would be happy to hear them and make corrections. This is admittedly an initial first cut at the magnitude of the stakes.

UPDATE: The RNC emergency stay application, filed the day after this post, significantly expanded the scope of absentee ballots potentially at stake. In light of that, be sure to see this updated estimate of the numbers at stake.

Share this: