Blockchain and voting: what can and cannot be verified

A voting system organizes how a community expresses a decision and how the result is verified. That community might be an association, participants in a protocol or a country’s electorate. Blockchain can record and verify certain parts of the process, with different applications and requirements in each setting.

Assessing its contribution requires distinguishing who can vote, how their choice is protected and how votes are counted. An accurate record is one part of that assessment: identity, ballot secrecy and protection against coercion also require safeguards.

What a shared register contributes

A voting system can use blockchain to record cryptographic commitments, publish results or check that events were included. This verifiability covers part of the process. It does not establish that a device displayed the correct option, that a voter acted freely or that identity was validated correctly.

Distinguish protocol governance, an association's consultation and public elections. Rights, voter registers, secrecy requirements and consequences of failure differ.

Problems consensus does not solve

• Identity: preventing duplicate votes requires a register and eligibility mechanism.
• Devices: malware can alter a choice before it is recorded.
• Coercion: remote voting may enable pressure or vote buying.
• Secrecy: pseudonymous addresses and public transactions may reveal relationships.
• Availability: network, interface or provider failures may exclude participants.

The National Academies of Sciences, Engineering, and Medicine explains that blockchain does not resolve fundamental internet voting problems. Resistance to changing a record does not equal end-to-end security.

What a design must demonstrate

Explain how votes are recorded, how inclusion is checked and how the tally is verified without revealing individual choices. Provide procedures for errors, appeals, recovery and independent auditing. Evaluate cryptography together with the interface, voter register and operations.

A technical trial does not establish suitability for binding elections. Applicable jurisdictional requirements and guarantees for the complete process need assessment. Avoid claims of impossible fraud or guaranteed anonymity.

From a consultation result to execution

Imagine a public governance vote on a protocol’s budget. The design must establish who may propose, when voting weight is measured, how delegation works and which threshold permits approval. These are system decisions, not automatic consequences of using tokens.

After voting closes, an approved proposal and an executed operation may still be different states. If a contract moves funds, review the exact instruction and who can stop or modify it. If a person implements the result off-chain, that operational dependency should remain explicit.

As an acceptance test, the team can reconstruct the result from the defined state and rules, compare it with the interface and verify the subsequent action. Rejected proposals, ties and an operator losing access also need testing. This example supports analysis of transparent governance; it does not establish that the same design preserves the secrecy required in a public election.

Separating eligibility, vote submission, counting and execution makes each guarantee’s source identifiable. A reproducible tally remains insufficient if the instructions eventually executed differ from what participants approved.

Where experimentation may fit

Protocol governance may accept public votes linked to addresses when participants understand the model. Even then, document power distribution, delegation, thresholds and the ability to modify rules. One token per vote does not mean one person per vote.

For a bounded pilot, define the threat model and success criteria first: eligibility, verifiability, accessibility, recovery and participation. Compare alternatives that already meet the organisation's needs. Results should establish which guarantees were achieved and which still depend on people or external systems.