[EM] danger of coercion (Re: First U.S. Scientific Election Audit...)

[Abd ul-Rahman Lomax]

At 03:22 PM 5/14/2007, Juho wrote:
> > And, something seems to be forgotten here. Elections are about
> > aggregating votes. Rarely do a few votes matter.
>
>Well, this matters at least to the individuals (and the mentality may
>escalate to wider circles too).

That's true. However, if there is a coercion problem that is 
spreading like this, it is a problem that surely will come to public 
attention and additional security can be put in place. It is fairly 
simple, for example, to prohibit all extraneous marks or signals on 
ballots. Write-ins can be handled with a separate procedure. Note 
that there is no way to avoid write-in coercion that is aimed at the 
write-in vote. "Write my name in or I breaka you face!" But very, 
very difficult for such coercion to actually elect this write-in candidate.

This is the point. Coercion, bad. But eliminating all possibility of 
a coercer knowing that the victim has complied is impossible. And 
keeping ballots private is not particularly effective.

What I'm saying is that (1) election coercion should be prosecuted 
and treated as a serious crime, as, indeed, should election fraud. It 
could be argued that election fraud is treason, because it is a 
betrayal of the sovereign, which in a democracy is the people, 
lawfully expressing their decisions through elections.

It is not a minor crime.

Why is it treated as such? Well, unfortunately, if election fraud 
attains its goal, the mechanisms of law enforcement are under the 
control of those who benefited from it. Only an awakened public can 
deal with a problem like this. Depending on the government to do it 
without serious general public support (to the point of a demand) is, 
quite simply, naive to the max.

Election coercion is not, from the point of view of the victim, the 
most serious form of coercion, by far. For one thing, it is fairly 
easy to comply, and the cost is low. So for a truly powerless victim, 
the danger can be avoided. *Of course* we don't want voters to be put 
in that position, but ballot imaging does not actually increase the 
risk seriously. Remember, in order to have any effect on attempts to 
coerce voting, the ballot must be distinguished in some way. This, 
all by itself, will leave traces that something unusual is happening, 
and the authorities -- if there is a public demanding it -- can 
actively investigate, not wait for a victim brave enough to come forward.

If the ballots are imaged, the public will know that they are being 
marked! If not, how will they know?

Discarding marked ballots is dangerous because it creates a ready 
method for those bent on election fraud to invalidate ballots. It can 
be done in a practically undetectable manner. Glue a piece of pencil 
lead to the tip of a finger. Pick up the ballot in a certain way, it 
is marked. Pick it up in any other way, not.

But if someone is marking ballots in this way, in numbers sufficient 
to affect elections, it will show on the images. Or there will be, if 
rules are in place to specially handle marked ballots, a large number 
of such ballots where images are not available. In any case the 
public will be able to tell something is amiss, if routine ballots are imaged.

The ballots should not be discarded. They should be counted under 
tighter rules about who can see them. They should be segregated for 
rapid special access if needed for audit.

My point is that the problem can be addressed far more effectively 
than by routinely keeping ballots out of public view. Remember, a 
coercer, under present law, can already arrange to view ballots 
directly. So how does routinely keeping them secret protect against coercion.

The protection is in the privacy of the voting booth and in measures 
intended to prevent connecting the ballot with the voter. This can 
only be defeated, generally, with the collusion of the voter, which 
is why the arguments about groups afraid to express their opinion 
because it would be unpopular is totally off the mark. Such people 
are not in any way put at risk by ballot imaging.

>An example of impact to bigger groups: Females are a majority out of
>which considerable part could feel the pressure of their husbands.

Some husbands might claim that their wives are a bigger risk....

I've never heard of a husband demanding that a wife vote in a certain 
way. I'm sure it has happened, all kinds of crazy things happen, but 
abusive husbands are usually concerned about more immediate and 
personal things than elections. How *dare* you disturb my beer bottle 
collection!

People with abusive spouses have a lot more to worry about than 
election coercion. Amongst all the serious problems, this one would 
hardly make a dent.

> > I think that standing the whole system on its head to avoid a very
> > theoretical and unlikely scenario is nuts.
>
>I think privacy in elections is a long standing healthy principle. No
>need to make radical changes. And if need arises, one can seek
>balance between different needs.

Privacy in voting is a long-standing tradition. There actually is no 
privacy right of the *ballots*. If I want to see them, I can go down 
and see them. Yes, I'll have to pay a city employee to stand there 
and watch, by the hour. But I should be able to find, amongst the 
ballots from a precinct, the one my wife cast, if she wrote in big 
letters on it, "This is my ballot, dear Hubby." Or she wrote my name, 
or a code name, in as a write-in candidate, more to the actual point.

If I'm willing to coerce my wife's vote, already a crime she could 
have me arrested for, surely I wouldn't mind spending no more than a 
couple of hundred dollars to verify that she complied or did not!

The whole line of argument is totally bogus. There is no material 
impact on voting privacy from ballot imagining. A narrow class of a 
narrow class might be affected, but in a way which would, in fact, 
make coercion easier to detect. And thus less of a problem, not more.

What I continue to find interesting is that people are aware of 
widespread problems with vote counting. Yet when something extremely 
simple is proposed that does not make any major changes in practices 
-- ballots are already public record, of a sort, and all that would 
be changing is that public viewing of them would become easy, instead 
of onerously difficult and thus confined to a small privileged group 
-- people line up to find things wrong with it.

It is one thing to anticipate problems, it is another to effectively 
discard the idea because one can think of a problem. Anything you do 
will have problems. But when we decide that something might be 
worthwhile, we don't toss it out just because we think of a problem; 
rather, we look at the problem itself. Is it real or is it an 
illusion? If it is real, can it be addressed? Will the proposed 
action make this problem worse or better?

(In the present instance, the possibility of coerced voting is raised 
as an objection, as if it were clear that this problem would be worse 
with public ballot imaging. But at least one line of thinking 
suggests that coerced voting is *already* a problem, though in most 
places in the U.S. and other stable democracies, not much of a 
problem, and that ballot imaging will make it *harder* to get away 
with coercion, particularly if attempted on a scale to tip elections.)

> >>  Open votes also are likely to lead
> >> to less votes to candidates that represent minorities and/or values
> >> that the voter does not want to reveal publicly. This could apply to
> >> minorities (political, ethnic, sexual, religious) or any deviation
> >> from the family, village, working place or country tradition and
> >> favoured values.
> >
> > Let me point out that in such an environment -- i.e., a minority
> > position is being hidden -- that minority position is unlikely to
> > win elections.
>
>Maybe in some two-party single-winner elections.

No. Not likely at all, except in multiwinner elections where a 
smaller fraction of the electorate can create a winner. You do not 
see near-majorities in the U.S. afraid to voice their opinion on the 
grounds that they would suffer harassment. But, once again, this is 
all beside the point. The issue of voting privacy is not relevant! 
Ballot imaging does not affect voting privacy, if the ballots do not 
routinely identify the voter. For it to be a problem *requires* the 
voter to act to make the ballot identifiable, which, quite simply, is 
not going to happen merely because the opinion being expressed on the 
ballot is unpopular. In a word, it is preposterous.

> > it has not been proposed that elections be public, only that
> > ballots be public.
>
>I took your term "open voting" to include also fully open processes
>and replied according to that assumption. (Also your Town Meeting
>example seemed to have such open processes. I also understood that in
>the "direct democracy power transfers" privacy of the voters could be
>limited.)

I mentioned open voting because such voting exists in the U.S., in 
Town Meeting. It also exists, by the way, in any elected assembly. 
The votes of elected representatives are not secret. In a small town, 
this is up close and personal. These people do not have police 
escorts, they go to work in the local store, they are completely 
exposed to the public.

I was pointing out that open voting exists and it does not seem to 
have the kind of problem that is being proposed as a serious one. 
When the scale gets large, problems could appear, because the 
absolute number of nut cases increases. But this is totally 
irrelevant to the issue of ballot imaging, I was only mentioning open 
voting because open voting is crucial to direct democracy. And 
arguing this here only expands the number of words without casting 
light on the issue of ballot imaging.

I have not seen, in this thread, one cogent argument against public 
ballot imaging, that stands up to scrutiny. The objection that ballot 
imaging will compromise voter privacy seems, like a number of 
knee-jerk responses we see in this field, to be reasonable, but when 
you look at *how* it might compromise this privacy, it turns out to 
have no substance.

(As another example of a knee-jerk response that seems reasonable but 
which disappears when examined, is the objection that Approval Voting 
violates the one-person, one-vote rule. And we only see that argument 
seriously advanced, now, by people who should know better, as part of 
political polemic, designed to win minds by confusing them. So we see 
FairVote trotting this one out, even though the same objection, 
fairly applied, would also consider IRV to be allowing voters more 
than one vote. Sauce for the goose is sauce for the gander, so 
ganders should be careful about selling goose sauce.)

> > In the situation described, a voter who feared that a vote would be
> > considered "deviant" simply would not make any marks to identify
> > the ballot. Why would he or she do this?
>
>In the case of coercion the cost could be e.g. one black eye. (And if
>the voting is public there's no need to even mark the ballot.)

The cost of what? The voter fears a black eye if the voter does not 
mark the ballot visibly?

Juho is arguing against public voting, when the context is clearly 
private voting, hence the mention of making identifying marks.

Public voting is an entirely different issue, and what I would 
recommend is, in fact, Asset Voting if we want to approach direct 
democracy on a large scale. Asset Voting creates electors who are 
public voters. Anyone can become an elector. And, yes, there would be 
safeguards against coercing small numbers of votes. Coercing large 
numbers of votes is actually not such a problem, because it is far 
more difficult to pull of without detection.

(And the *real* problem: lack of prosecution of election fraud of all 
kinds, coercion being only one, and probably not the most serious of 
the real problems. It's nasty for the one coerced, but, frankly, if I 
was gong to be coerced, I'd rather be coerced into voting a bad vote, 
nasty as it might taste, than into much more serious losses. The 
value of a vote is probably on the order of a few dollars, certainly 
not a few thousand dollars.) (The proof of the value of a vote is 
what people are willing to spend merely to influence them, not to buy 
them. And there are, what, 80 million voters in the U.S. presidential 
elections, and the total spending is what, on the order of $800 
million? That's ten dollars per voter.)

> > But, in any case, this is moot. We are not proposing public voting,
> > only that public voting does exist and does not seem to have the
> > level of problem that is being asserted.
>
>What does word "we" refer to?

All of us, except anyone who is, here, proposing general public 
voting. I didn't. Did you? Did anyone?

I *described* public voting that already exists. Surely if there is a 
coercion hazard, it is stronger with full public voting than with 
mere ballot imaging!


> >> Therefore secret ballots are a good main rule (exceptions allowed but
> >> justification needed).
> >
> > Secret ballot can be appropriate for elections; but I would reverse
> > what was said here. Secret ballot, in a mature system, would be the
> > exception, not the rule.
>
>The emergence of that level of society might take time. I think for
>the coming years secrecy might still be the best main rule in normal
>public elections.

And, again, nobody here has proposed differently. Nor would I even 
propose, even in a mature system, that "elections" involve open votes 
at the base level. Only the votes of electors would be public.

That's the U.S. Presidential system, by the way. Electors vote publicly.

> > It's quite difficult to corrupt direct democracy. Once the
> > transfers of power happen secretly, it becomes easier to corrupt.
>
>These words seem to indicate that in direct democracy we would need
>to seek some balance between privacy and risk of corruption.

Yes. When I referred to open voting above, I was referring to 
small-scale direct democracy. A small unit within a large city, a 
precinct, perhaps, can have a precinct council at which any citizen 
can appear and vote.

I would make these councils more fully democratic by allowing 
citizens to be represented in them by proxy, and I would use 
delegable proxy to make it far more likely that every citizen would 
actually be represented, who cared at all to be so.

And to me, it is the process that is far more important than the 
*votes*. Being represented is the issue, to me.

On a larger scale, steps must be taken to restrict floor access, or 
meetings become impossibly difficult. But that is another problem.

Where there is any fear of coercion as a problem, which seems to be 
the case now (that there is a fear), then Asset Voting feeds the 
delegable proxy system, allowing secret assignment of votes under 
secure and identity-validated conditions. If you have an Asset Voting 
system, by the way, the electors don't cast their own vote, they only 
cast votes that they gathered in the polls, and they vote themselves 
in those polls. So their vote is assigned like everyone elses. They 
can name themselves or not.... If they want to insure that they have 
at least one vote, they simply vote for themselves....

Under difficult conditions, I would filter out "candidates" who 
receive less than a certain number of votes. These candidates would 
be contacted by authorities and given an opportunity to reassign 
their votes, in private, to create electors who hold enough votes 
that the public can afford ot offer them protection. The number 
depends on conditions. In small towns, my opinion, the minimum number 
of votes could be very small.

(Voters getting under the minimum would not be told how many votes 
they had received. I'll leave it as a puzzle, why. They would revote 
and their vote would assign all their votes, but how many votes that 
was would not be disclosed to them. All they would know would be that 
the number would be under, say, five.)

And this is beside the point and has nothing to do with ballot imaging.

Ballot imaging is a concrete and simple proposal, cost-free (to the 
public, costs would be covered voluntarily by those interested in 
obtaining the images), that should practically nail down the issue of 
public confidence in elections, and it is vastly superior and far 
less corruptible than all the complex voting machinery that has 
become so controversial. Further, it has so many other beneficial 
side effects, we can anticipate, that there is no doubt in my mind 
but that we should be going for this as soon as possible.

It could be done practically immediately. There is no equipment to be 
purchased. There need be no appropriations. It *would* take hearings, 
I'm sure, to hash out possible problems, but as far as I can see, 
those problems would be few and easily soluble.

Why am I so interested in FA/DP? Because the Delegable Proxy 
structure, among other things, provides a ready avenue for ideas to 
be easily and efficiently considered, and to receive *responses*. If 
we had DP in place, I would know, quickly, if this imaging idea were 
going to be accepted or not. Or at least if it were going to be 
*considered* or not!