This is the 4th of a 7-part series of election security vignettes intentionally kept as brief as possible to stick to the main point of the title above. Its intended to acquaint relatively new-comers to election integrity and security, and that includes anyone and everyone from concerned citizens, to journalists, to policy makers.
In Part 3, I hope I made a strong case that there is only one main reason why we have to move away from having to place complete trust in computers to count ballots right – you can’t trust computers! And hence we don’t really know if computer-generated election results are correct. Every other concern is secondary.
However, I got some push-back on the not trusting computers thing, so let’s break it down to basics. To start, let’s try to tie the issue to daily life with computers.
Example 1: playing computer games. It’s a drag if you lose a game “unfairly” because the game software made a mistake; or if your game evaporates because the software crashed your computer, or your phone ran out of power. But did you lose anything of huge value? For many people, no, and though vexing, it is not a surprise. Stuff glitches out all the time; it happens.
Example 2: similar situation, but text processing: something goes wrong, you lose some important work, and you have suffered a real loss. And so you take precautions like saving your work, doing file backups, etc. But when you’re finished, are you concerned that software bugs put in words you didn’t write? Maybe not much, but you have a simple protection that you do anyway: you proofread your work before committing it to paper or to digital delivery. If you don’t, then some embarrassing auto-correct might leave you with a red face. So you check the computer’s work.
Example 3: banking. Losing money, now that’s something most agree is a real loss. With banks and credit cards it is “trust but verify” for many people – checking a statement or ledger to make sure that each time some of you money went elsewhere, it was you, rather than a bug in the bank’s software, or human error of a bank employee that the bank didn’t catch, or some credit card fraud or e-banking fraud. You have something valuable at stake, so you do an “audit” – that is, you check up on what all the computers have been doing with your money, to the best of you ability. And if it’s not you, but big business, they do an “Audit” with a capital A, and with professional accountants.
Back to Voting…
With voting and ballots, we use computers, and we want to do a similar “audit” to detect and correct mistakes that the vote counting computers may have made. But to get back to the basic issue, why can’t we trust that the voting counting computers are doing a good job? Why don’t we trust claims that voting systems are “secure” or “bulletproof” or “best quality” and so on?
The basic answer is that all voting technology have computers inside, and they are no different in fundamentals than any other computer on your desk or in your pocket. You’ve seen computers let you down before, so don’t assume that your election officials’ computers will always work perfectly. The bad news is that they won’t always work perfectly.
The good news is that it is easy to detect and correct when ballot counting computers make mistakes: spot check their work with human examination of sampled paper ballots. And if the spot checks spell trouble in bigger and bigger batches, there is always the recourse to hand counting all the paper ballots … but only if there is one paper ballot per voter! Which is why we need to stamp out paperless voting! (That also goes for its close cousin, old “DRE” machines that produce crummy little thermal paper tapes that it turns out few voters ever look at and nobody believes is a substitute for a real ballot).
And I believe that that will happen quicker, once more people understand that voting systems are just computers, and you shouldn’t blindly trust your vote to a computer that, like all of them, can make mistakes.