A ballot’s job isn’t finished the moment it’s cast. From the second a voter drops it in the box to the moment a final result is certified, that ballot has to survive a long chain of handoffs — sealed containers, transport vehicles, storage facilities, counting boards — and every link in that chain is a place where trust can either be reinforced or quietly eroded.
We use the term “chain of custody” deliberately, borrowing language more commonly associated with evidence handling in a criminal case. The comparison isn’t accidental. In both contexts, the goal is the same: being able to demonstrate, step by step, that the thing in front of you now is the same thing that started the process, untampered and accounted for the entire way.
In practice, this means tamper-evident seals on ballot containers, logged signatures every time custody changes hands, bipartisan teams present for transport and storage whenever possible, and video monitoring of secure storage areas. None of these measures are exotic. They’re closer to standard inventory-control practices, applied to something with far higher stakes than a warehouse shipment.
Where chain-of-custody procedures tend to break down isn’t usually in the big, visible moments — election night itself is typically well-staffed and closely watched. It’s in the quieter intervals: overnight storage, the days between polling and certification, and any point where ballots are moved between facilities. A gap in documentation during one of these windows doesn’t necessarily mean anything went wrong, but it does create a gap that’s hard to fill retroactively, and that gap is exactly what erodes confidence even when the underlying count was accurate.
This is part of why we keep returning to the importance of pairing strong custody procedures with the kind of post-election audit we wrote about previously. An audit is only as trustworthy as the ballots it’s checking. If custody has been airtight, an audit can confirm that with real confidence. If custody has gaps, even a clean audit result leaves room for doubt, because there’s no way to fully rule out that something changed between casting and counting.
We’d also flag that chain-of-custody standards vary considerably by state and even by county, which makes this an area where relatively modest, low-cost reforms — standardized seal logs, mandatory bipartisan witness requirements, published transport schedules — could meaningfully raise the floor nationwide without requiring new technology or major new spending.
None of this is about assuming bad faith on the part of election workers, the overwhelming majority of whom take their responsibilities seriously. It’s about building a system that doesn’t have to rely on anyone’s word alone — one where the documentation itself can answer the question “how do we know nothing changed?” before anyone has to ask it.