This is an idea from a post previously on this website but before it got wiped out. Here I will be rehashing this.

Prior to the 2024 election, a few states revealed the early voting data prior to close of polls. Some of this data was quite helpful in making predictions about the outcome of the election. Because the data was shown prior to close of polls, it's plausible that it had a slight influence on the result of the election. The results of the election having an impact on the result of the election is not great.

One way we could prevent this from happening in the future is a similar disclosure model to how people online make predictions that don't affect the outcome. They have their prediction in text, and they compute the hash of it. Then they post the hash publicly. Once the time frame in which the prediction can no longer have an impact on the result, a person would then reveal what the original statement was. Now others can use that original statement to compute the same hash. This validates that the person did in fact write that statement prior to the event taking place.

Rather than use this technique for predictions, we can use this sort of thing for certifying the authenticity of data prior to the close of polls but without actually revealing any of it. The early vote data could be disclosed as a series of hashes over time. Once the election has finished, the data can then be revealed. The public can use that data to validate the original hashes posted prior to the election.

Another option would be to simply not disclose the early vote data until after the election. While this solves our original problem of preventing the data from impacting the election, it has another undesirable aspect. If a government stops releasing any kind of data related to an election to the public, it undermines trust in elections. This hashed disclosure method eventually gets all the data into the hands of the citizens. They can at least verify that at the time the data said a particular value. This should sate concerns about post hoc tampering.