May 29, 2017

Absentee Voting No Panacea

Various groups that oppose paperless electronic voting have recommended an alternative: if you really want to be sure your vote is counted, vote absentee. Having studied e-voting, and living in a county with paperless e-voting, I sympathize with the desire for an alternative. But it should be noted that absentee voting offers iffy security as well.

The best alternative to risky e-voting, where feasible, is in-person voting with a paper ballot. This allows the main election safeguard, which is the presence of observers from diverse political parties, to operate: the observers can watch the voter check-in process, watch the ballot boxes, and watch the counting of ballots.

With absentee voting, by contrast, the distribution, validation, custody, and counting of ballots are generally less transparent. You’re pretty much stuck trusting the country clerk and his/her staff. I don’t want to cast aspersions on the virtue of any particular county clerk; but I hope you’ll agree that a more transparent system is better.

Absentee balloting also weakens the secret-ballot guarantee, which requires that nobody can verify how another person voted. This is an important safeguard against vote-buying and intimidation, as it frustrates the vote-buyer’s or intimidator’s ability to know that he got his way. Absentee voting allows the voter to prove to somebody else (say an employer, union boss, or abusive spouse) that the vote was cast a particular way. That’s a serious drawback.

Now it may be that you don’t want to sell your vote, you don’t fear intimidation, and you trust your county clerk. For you, absentee voting might be the best available substitute for e-voting in person. But encouraging widespread absentee voting is not a good public policy response to the e-voting problem.


  1. Stephen Cochran says:

    I think a sudden surge of people voting absentee instead of using “recordless” voting machines will do a world of good by educating the election officials of just how little we trust the new systems.

    Imagine what would happen at the county clerks’ offices if 10% of the voters went for absentee ballots, nationwide. It would make the outcry over Florida look like a slow news day.

  2. Hal Finney says:

    I agree with your cautions about absentee voting. I think the security community has a tendency to go too far in our critiques of electronic voting and ignore the problems in existing systems. We know the failure modes of computer systems so well, they are obvious and glaring to us. We criticize registrars for thinking that computers are a sort of magic security fairy dust which will solve their problems. We are painfully aware of just how far they are from this ideal.

    But we fall into the same trap, thinking that paper provides its own magical security features, while the registrars are the ones who know from long and painful experience just how insecure paper based voting systems are.

    The most remarkable feature of the current debate is how the two communities are talking past each other, unable to hear the points the other side is making. All each side sees are the problems with the technology that they know the most about. No one seems to be in a position to make an unbiased evaluation of the relative merits of the various voting technologies that exist.

  3. Just and oddity: Note that West Virginia doesn’t require secret voting… a voter may choose to keep their vote secret or not.

  4. Chris Tunnell says:


    What happens if they don’t keep their vote secret?

  5. Jordan Vance says:

    So here is a pretty easy, if expensive, solution. Have the voting machine print out an enumerated list of who you have voted for. You then go put this in a box, and if a recount is needed, they can simply go to the paper trail. Sure, it might take a while to recount, but at least it will be right.

  6. Hal Finney says:

    Jordan, what if the paper-counted votes don’t match the ones which were recorded electronically? Why would you assume that the paper count was more accurate? Don’t you need to estimate an error rate for the two systems?

  7. Jordan Vance says:


    Yes, you never assume that something will measure the vote perfectly. But if the machine has a screen that says you vote for Candidate Y, asks you to confirm this, and then prints a piece of paper that says 1 Vote for candidate Y, then there are a lot of checks to make sure that the paper vote receipt confirms the electronic vote. You are correct that we can’t assume the paper count is more accurate, but it is possible to say “Hmmm…” if there is a huge disparity. What to do after the count is done is, well, unknown to me, but I do see where the possibility occurs that someone tampers with the printouts to get a machine that recorded the votes correctly tossed from the election. But it is slightly more transparent to everyone.

    Well, here’s a scenario… say the e-machine at Polling Place X gets changed to swap every 6th vote for a candidate to his opponent. So if 600 people vote, and there is a 200 vote difference between the paper ballot and the electronic count, something is fishy there. Furthermore, it’s one thing if say .01% of people say the wrong name is on this print out. If 1% or 10% or .25% complain about having the wrong name on their printouts, then it is something to look into. This isn’t some local election we are talking about. It’s been 4 years, the system still sucks, and there shouldn’t be any reason to have as many checks on the systems as possible.

    In other news, my sister’s wedding is this weekend in Jacksonville. I’m hoping that the storm misses the US completely, but I’m really praying that it doesn’t ruin her wedding.