There has been a lot of arguing about vote-by-mail these past few days, and I think the main objection to VBM has not been completely understood.
For clarity's sake here it is again, before I back it up:
We should never accept a voting system that lets a woman's husband watch as she fills out her ballot.
People have belittled this objection by exaggerating it. They picture some psycho controlling husband who will force his wife to vote (R)---and then they argue that this problem is minor in terms of raw numbers. How many psycho controlling husbands are there?
But this is a big fat straw man. The problem here is not confined to abusive relationships. Even in a perfectly "normal" marriage, of which there are tens of millions, people will defect from (R) to (D) one year and keep that secret from their spouses. This is a widespread phenomenon, large enough to swing elections.
Below the fold, some evidence that yes this is a huge problem. Plus some counter-arguments dismantled.
I will say it again:
mandatory ballot privacy protects a voting bloc ("secret defectors") comprising millions of voters, large enough to swing national elections. This is not an extreme case of a few psycho husbands.
For example, my own parents are not abusive at all. They vote (R) --- but my mother is more sympathetic about abortion. I can guarantee you that if she decided to vote (D) one year she'd keep it to herself. If she couldn't keep it to herself she wouldn't defect. This is without any fear of violent reprisal, without any abuse, without any psycho controlling husband.
Now what causes that conformism, if there is no overt threat? Simple: people behave according to local mores when they are being watched. Rush is on the radio every morning, and Dems are routinely described as idiots and traitors, so you have local mores. You don't need an abusive relationship for this to happen.
Some basic numbers
How many secret defectors are there? This is difficult to measure, but we have some tantalizing clues. Observe this statistic, which admittedly consists of suspiciously round numbers:
Over 75 percent of married men were certain that their wives voted along their family's party line. When asked the same question, only 50 percent of the women responded the same way.
Let's do some math: if these numbers are accurate, the smallest possible percentage of secret defectors is 25%. (We might foolishly multiply 75% by 0.5, but these two statistics are not likely to be independent.) 25% of approximately 55 million American married couples means about 13 million married women who secretly do not vote as their husbands think. Thirteen million is a lot of goddamned people. This is a big problem, QED.
Even if we employ the we-don't-know-shit principle and take this down by an order of magnitude to be safe, you're still talking about a macroscopic percentage of American married couples. Enough to decide who wins in any swing state.
In short, don't give me this bullshit argument that we're talking about some rare or freakish phenomenon. There is a huge voting bloc here, whose choices are protected by the default security of a ballot booth.
Common myths exploded
Let me address the remaining counter-arguments I have heard about VBM. I will enumerate them for your convenience:
- "Oregon has decades of experience with it, and it works."
There is no doubt that OR residents really like the system, that it is convenient, cheap, and easy. But that only means it works logistically. This is not evidence that a security hole isn't a problem.
In fact you can say many similar things about touch-screen voting machines. They reduce the undervote, a lot of people prefer them in terms of convenience and ease-of-use. But of course convenience is not the problem.
- "The people who are most prone to intimidation like it too."
To my knowledge, nobody has ever asked the specific group who would be disenfranchized---namely, secret defectors. Does anyone know of a poll in which people were asked if they planned to defect, but had difficulty arranging that? And if those people like VBM?
- "If anyone intimidates a family member, then he'll go to jail."
We are not talking about an actual act of intimidation. There is nothing to arrest anyone for. See my example above. You can't arrest my Dad for listening to Rush every morning, or projecting a general air of contempt for Democrats.
- "What about all the abuses with electronic voting machines?"
What about them? You aren't forced to choose between Diebold and VBM. You aren't forced to embrace one security problem to cure others.
Indeed, this national VBM idea may be a repeat of the 2000 fiasco. We embraced those damned machines because we were sick of butterfly ballots and chads. So to solve those problems we rushed into another voting system that sacrificed the basic security of free elections. Wasn't that a bad idea? So why would we solve our new voting problems by throwing away yet another fundamental tenet of secure elections?
- "People could watch me vote on the touch-screen machine too"
That's because the touch-screen machines are so insecure that you can't put them behind a curtain! If anyone can subvert an AccuVote in 60 seconds, then it has to be in view at all times.
But this only a problem with DRE machines. Me, I voted behind a curtain. In fact, I have always voted behind curtains, both with punch paper ballots and switch machines. This is the minimum level of privacy I expect, and it is easy to arrange.
- "People can vote away from their husbands/families if they want."
...and then they give away that they are defecting. No dice. One reason why we have ballot booth privacy by default is that privacy doesn't look suspicious. If you must opt for privacy, then you project your intent.
- "But we already have absentee ballots; don't they have the same problems?"
Yes, but they are not the default. In our electoral system privacy is opt-out, not opt-in. Opt-in privacy is dangerous because it projects your intent to make an upopular choice.
But yes, absentee ballots do have this security problem. They are still a minor problem because you have to go to the trouble to arrange that scenario; it is not dropped on every family's kitchen table by default.
- "If anyone is in that situation, she should wonder why she married an asshole."
Believe it or not, I have heard this argument a half-dozen times this week. From progressives.
It is not worth elaborating on why this is a stupid remark---but I will point out again that we are not just talking about abusive relationships. People need secrecy to defect even in "normal" households.
- "There are all sorts of logistical problems with voting by ballot box."
Then please, solve this one deal-breaker problem with VBM. Design a VBM system that doesn't have this problem. If VBM is so much better (and I believe you when you say it is cheaper/safer/easier/popular,) then you should have a strong incentive to fix its Achilles heel.
- "Well, then what do you suggest?"
I suggest you ask the scientific and crypto community. There is a lot of ongoing research in the development of secure election protocols, with the goal of solving our existing problems without creating new ones.
You ask anyone developing election protocols today---or ask anyone who breaks them, like the Princeton people who hacked the AccuVote machine---and I guarantee you that none of them will agree to a voting system without ballot booth-level privacy.