The Cure Period Experiment
In Chester County, Pennsylvania, election workers this past November processed 47,000 mail-in ballots with what they call “deficiencies” — missing signatures, incomplete dates, or mismatched information. Under Pennsylvania’s relatively new “cure” process, voters had until 5 p.m. on the Friday after Election Day to fix these problems. The county contacted each voter by phone, email, or mail, and ultimately saved about 31,000 votes that would have been discarded entirely just four years ago.
This administrative detail might sound boring, but it’s actually a huge expansion of voting access in recent Pennsylvania history. It shows why electoral mechanics matter way more than most people realize. Whether your ballot gets counted often comes down to procedural rules that vary wildly across states, counties, and even individual precincts.
How Small Rules Create Big Problems
Pennsylvania’s experience shows how seemingly minor regulations create major gaps in voting access. Take signature matching, which exists in some form across 31 states. In Allegheny County, trained election workers compare mail-in ballot signatures to voter registration records using a multi-step verification process. If there’s a questionable match, three different officials must review it before rejection. But some Pennsylvania counties rely on temporary workers with minimal training, leading to rejection rates that vary by as much as 400% between jurisdictions.
The timing rules add another layer of mess. Pennsylvania allows mail-in ballots postmarked by Election Day to arrive up to three days later, but only if they lack a clear postmark. That rule emerged from 2020 litigation and applies differently depending on postal service efficiency in each region. Meanwhile, Georgia requires mail-in ballots to arrive by 7 p.m. on Election Day regardless of postmark, and Texas demands they arrive the day before. These variations mean that an identical voter behavior (mailing a ballot on the Saturday before Election Day) produces different outcomes depending solely on geography.
The cumulative impact is huge. Research by MIT’s Election Lab found that ballot rejection rates in the 2022 midterms ranged from 0.03% in Vermont to 3.1% in Arkansas. Procedural differences accounted for roughly 60% of this variation after controlling for demographic factors.
The Infrastructure Question
Behind every voting rule lies a practical question: who pays for implementation, and how much does it cost? Pennsylvania’s cure period requires counties to contact thousands of voters within a compressed timeframe, verify their responses, and update records accordingly. Philadelphia County estimates this process costs approximately $127 per cured ballot when accounting for staff time, communication expenses, and system updates. Multiply that across the state’s 67 counties, and cure periods represent a multi-million dollar investment in election infrastructure.
This cost calculation matters because election administration remains primarily a local responsibility, even for federal contests. Rural counties with limited budgets face different tradeoffs than urban areas with larger tax bases. Lackawanna County, for instance, struggled to implement robust signature verification because hiring additional temporary workers would have consumed 23% of their entire election budget. They opted instead for a streamlined process that reviews only signatures flagged by automated systems. It’s a reasonable compromise, but it means voters get treated differently than they would in wealthier jurisdictions.
The technology gap makes these disparities worse. Montgomery County uses optical scanners that can process 300 mail-in ballots per hour and automatically flag potential signature mismatches for human review. Cameron County still relies on hand-counting that processes roughly 50 ballots per hour with no automated verification systems. Both methods can be accurate, but they create different timelines, costs, and error rates that ultimately affect whether votes count.
Federal Constraints and State Innovation
The constitutional framework that governs American elections creates a strange dynamic: states control most voting procedures, but federal law sets minimum standards that occasionally conflict with state innovations. The Help America Vote Act of 2002 requires provisional ballots for voters whose eligibility is questioned, but it doesn’t specify how states should verify those ballots or what timeline they should follow. Pennsylvania interprets this to mean provisional ballots can be counted up to seven days after Election Day if documentation arrives. Ohio sets a four-day limit. Nevada allows ten days but only for military and overseas ballots.
This federal-state tension becomes most visible around voter registration deadlines. Pennsylvania allows same-day registration during the early voting period but requires voters to register 15 days before Election Day to vote by mail. That distinction made sense when mail voting was limited but creates confusion now that no-excuse mail voting is available statewide. The Voting Rights Act requires changes to voting procedures in certain jurisdictions to receive federal preclearance, but Pennsylvania is not a covered state, leaving these policy choices entirely to state discretion.
The result is what election law scholar Rick Hasen calls “electoral federalism chaos.” We have a system where fundamental voting experiences vary not because of principled policy differences, but because of overlapping jurisdictions, funding disparities, and historical accidents.
Measuring What Works
Figuring out how to evaluate electoral systems is surprisingly tricky. Ballot rejection rates provide one metric, but lower rejection rates don’t automatically indicate better systems. They might reflect weaker verification processes. Voter turnout offers another measure, but higher turnout could result from demographic changes rather than procedural improvements. Pennsylvania’s experience suggests that multiple metrics matter: ballot accessibility, election security, administrative efficiency, and public confidence.
Early data from Pennsylvania’s cure period looks promising across several dimensions. Cured ballots show the same demographic distribution as successfully submitted ballots, suggesting the process doesn’t systematically advantage particular groups. Post-election audits in counties with robust cure processes show error rates consistent with traditional ballot processing. Perhaps most importantly, post-election surveys indicate that voters who used the cure process report higher confidence in election integrity than those who had ballots rejected in previous cycles.
These Pennsylvania experiments matter beyond state borders because they generate evidence about what works under real-world conditions. Unlike academic simulations or small-scale pilots, statewide implementation reveals how electoral innovations perform under pressure, with limited resources, and amid political scrutiny.
The Laboratory Problem
The deeper challenge is that electoral experimentation creates ongoing uncertainty for voters and administrators alike. Pennsylvania has changed its mail-in ballot procedures four times since 2020, each adjustment requiring new voter education, staff training, and system modifications. This is responsible policy development in one sense, adjusting rules based on evidence and experience. But it also means that election administration exists in a state of perpetual transition, making it difficult for anyone to develop expertise or confidence in the system.
The evidence suggests that Pennsylvania’s electoral innovations generally improve ballot access without compromising security. But evidence alone doesn’t resolve the fundamental question of how much variation in voting experiences American democracy should tolerate, or whether federal standards should override state experimentation in service of greater uniformity. These remain political choices that technical analysis can inform but cannot resolve.
