Once More About Elections Processes and Fraud

If you haven't yet listened to the challenges to the election results, you should.

When I posted about usually getting the president we deserve, I was assuming that none of the factions in the current contest would be stupid enough to make use this soon of the tools of voter fraud that they have been putting into place. I was naively hoping that there would be time for common sense to reign back the erosion on election practices that has taken place.

I had been hearing what the normal news outlets have been mentioning about the challenges that the Trump Campaign legal team have been mounting with a mixture of, I don't know, cynicism, ennui, exasperation, disappointment, rolling eyeballs, etc., that they would be "dragging this out yet again".

Last night, a friend posted a link to a youtube video of the Trump campaign's legal team' press conference:

https://youtu.be/buQCdCSDWQQ

I don't know why I find it astonishing. I don't know why I want to be so naive.

Voting has to be a transparent process to be valid. 

In the simplest case, where voters are not intimidated by threats from the parties presenting their proposals and candidates, secrecy is not necessary. A simple raise of the hands in favor or against is all that is necessary. Everybody can see who voted for what/whom, and everybody can count.

But people get their ego's all tangled up in the results of elections, and then attempt to alter the outcome unnaturally. So the simplest case doesn't scale well.

There are many ways to try to alter the results of elections:

• Campaigning itself is one such attempt.
  
• If campaigning doesn't appear to be working, and people aren't willing to accept that, they might use intimidation, explicit threat, and actual force to prejudice the outcome. 
• And if that doesn't work, they might prepare to attempt to alter the outcome by legal technicality after the fact. 
• But legal technicality relies on vagaries of courtroom process, so, if, during the election, it becomes apparent that they are going to lose, they might try to alter the outcome by interfering with the process, so they don't have to gamble on those vagaries.
  

Election best practices has provided a means of circumventing these problems.

Elections best practices require that the ballot be cast in secret, but counted in the open.

With or without a lot of thought, it's clear that there are contradictory requirements here: cast in secret, count in the open. 

Once it's open, it's no longer secret.

Somehow, you have to reliably separate the identity of the voter with the content of the vote in between the moment the ballot is cast and the moment it is counted.

Here is one simple way to do it: 

1. A place to store the ballots which have been cast -- a ballot box -- is prepared, and inspected and shown to be empty, and secured before voting starts.
2. The ballots are counted before voting starts, to be sure there will be enough for all the registered voters, and to be able to check the number of ballots in the box and the number left over against the number of voters receiving ballots.
   
3. Judges, ballot handlers, and observers are also provided. Observers must include representatives from all parties with interest in the results, and they must be allowed to actually stop the process and ask for corrections to be taken, if irregularities are observed.
4. Votes are taken as follows:
   1. A voter requests a ballot.
   2. The voter is given a ballot and some sort of covering that hides the content of the ballot. Neither the ballot nor the cover provide a means of identifying the voter.
      
   3. The voter is provided a place to mark the ballot in private.
   4. The voter marks the ballot by hand in private,  and places the ballot in its cover before submitting it.
   5. The ballot is submitted directly to the election judges, who transfer it directly to the ballot box without exposing the ballot contents.
5. At the end of the designated balloting period, the ballot box is opened and the contents counted in the presence of the judges and observers. The counting process itself has to be observable. If machines are used, they should only be used to verify a hand count.
6. Judges and observers must be allowed to record the counts taken and take their records with them.
7. The ballots must be packaged and the packages sealed before being transported from the balloting place.
   
8. Ballots and counts must be transported to a central place where the ballots can be securely stored until the election results have been properly certified, and the counts can be tallied with the results from other ballot places, and the results for each balloting place must be published.

The reasons for being so particular are roughly as follows:

1. If we aren't careful with the ballot box, it's too easy to stuff it with fake votes.
2. Counting the ballots used and the ballots not used is a way to anonymously check that the box hasn't been stuffed and that valid ballots haven't been discarded.
   
3. Without observation and without someone competent to judge, it is way too easy for the election place staff to do all sorts of things to undermine the results. Especially, observers from each party can help to keep the others honest.
   
4. Concerning what the voter does:
   1. It's way too easy to buy the vote of someone with an unrequested ballot.
      
   2. Disclosure of the contents of the ballot also subjects a voter to potential reward and/or retribution.
      
   3. Observation of the voting being cast is another way to discover the contents of a ballot.
      
   4. Use of a machine provides places to hide devices to eavesdrop on the vote. In fact, it's hard to design an electronic machine that would not leak at least some of the details through radio noise.
      
   5. The more hands and the longer a distance a ballot passes through before being put in the ballot box, the more opportunities there are to use sleight-of-hand to misdirect the ballot, slip an added ballot in, or surreptitiously observe the contents before the ballot goes in the box.
5. Counting the votes once at the voting place helps assure that the ballots that are sent to central storage are the same as the ballots that are received there. It means that more people have to be present from the time the polls close until the counting ends, but that is a good thing. More eyes reduces temptations and provides more opportunities to blow a whistle on fraudulent activities. Moreover, no voting place should be set up to take more votes than two or three people can count and check by hand in an hour or two. And machines at this stage are too much temptation for hidden shenanigans.
   
6. The more records of the results at the various stages, the more opportunities to confirm their validity.
7. Sealed packages are significantly harder to open and alter the contents of than unsealed packages.
   
8. Central counting is too much temptation for shenanigans, if it's the first count. On the other hand, centralizing the second count allows the use of machines at that point. Secure storage and published results from each step allows greater confidence of the results, and of the results if a review and recount is required.
   

• First observation: Yes. This takes time. 

There is no necessity that the process be finished overnight, or even in a day. This can actually a good thing, because it can help prevent parties with too much interest in the results from knowing whether they want to take the risk of interfering with the process until it's over.

• Second observation: Machines which can be used to hide or expose parts of the process must not be used in the voting or counting process.
  

Except, there might be limited use for those voters whose physical limits would prevent them from marking the ballots by hand. And they can be used at a central location to confirm the original hand counts.

• Third observation: Mail-in ballots prevent the use of both observers and judges in critical parts of this process. Their use must be limited to by-necessity-only.
  

It's extremely frustrating for me to have to point this out, but it's just not possible for judges and observers to see what happens between the time the ballot is requested and the time it is submitted, to make sure that the ballot does not get diverted to fraudulent purposes. 

And it's not easy for them to observe the request and submission process, either.

Cries of "But mail-in is so much less stressful!" notwithstanding, mail-in, if allowed, must be limited to cases of necessity.

• Fourth observation: Society must support the polling process in ways we do not presently support it, if we want valid results.
  

People whose jobs prevent them from attending a regular voting place during the designated times are prevented from voting, and that also biases the results.

But mail-in is not a solution.

This is a place where we can improve current practices. I can suggest a few things that would help:

1. Employers should be required to give employees necessary paid time off to go vote.
2. Voting judge, observer, and counting duties are no less important than jury duty. Failing to properly staff a voting place is just begging for fraud to occur. We should be willing to do something similar to jury duty for election duty.
   
3. Alternative voting places and times should be provided instead of mail-in. Alternative voting places would make it easier to provide election judges and observers, and more possible to confirm who has already voted. Computer systems could also legitimately be used to determine who has voted already, but only carefully secured systems.
   
4. Mail-in ballots, if used, must never be opened and counted until after the voting places have closed and all voters have been confirmed, to avoid encouraging ambitious voters from doubling up. They must also be stored securely and opened and counted in the presence of judges and observers, just as in-person ballots.
   

• Fifth observation: Counting votes by machine to confirm the original hand count is an extremely simple process. It does not require the use of software developed by big companies, and especially not by foreign companies. 

I suppose I should expand a little on my impressions of the charges being made.

People working for the election are supposed to be working under oath, or as if under oath. Trump's legal team has found people willing to testify under penalty of perjury that people who are working for the election have broken their oaths.

Should I believe a hundred or so odd-ball malcontents who are willing to cooperate with the monkey pretend president with the orange hair? Or should I believe the thousands and hundreds of thousands of dedicated election workers?

I know how easy it is to get swept up into following the egoist.

And I have also worked real jobs for a long time. I know how easy it is for dedicated, hard-working, well-intentioned workers to say, 

This is unreasonably difficult. Getting the job done is more important than getting it done right!

and carelessly invoke truisms like

The perfect is the enemy of the good!

and convince each other to do things that shouldn't oughta be done.

I don't know who has broken their oaths. But it is apparent that a number of people have. At bare minimum, a number of people are seriously deluding themselves about what is right. 

I hope we can solve this without somebody starting witch hunts, but it looks difficult, unless one side yields things they think they shouldn't yield, which is also a bad result.

Whatever happens in these challenges to the election results, we should be looking at these as a wake-up call. We shouldn't be so caught up in our pursuit of good ends that we ignore the potential damage from our methods. 

We are treating our elections processes way too lightly.

Revisiting the Vote -- Voter Fraud

(This is a bit of a different way of thinking about vote fraud.)

I presented An Awkward Proposal for an Amendment to Correct Election Processes some time back.

Re-reading it now, I can find numerous holes in it. Some day I'll re-work the proposal, but today I want to think about voter fraud.

At this point in time, both the major US political parties are accusing each other of voter fraud. Dead voters, multiple voters, non-citizen voters, influencing absentee ballots, cutting voter districts to water down the opposition, ..., ..., ...

Thinking about ways to protect against voter fraud, I realize that most protections would have results worse than the fraud itself.

That is, being a computer scientist, I tend to quickly think of the vulnerabilities induced by electronic voting and by improper use of absentee ballots and such. I also tend to think about safeguarding the processes as they exist, and those tendencies are reflected in the post I link above.

But that does not really get at the source of the problem.

In order to get into the proper frame of mind to consider what to do about voter fraud, we should start with a realization of something I will point out in a few paragraphs. To get there, we should start with an understanding of what a voting process is.

Many times, it is described as a way to get a consensus of opinion from a body of people.

But what people mean by "consensus of opinion" varies widely.

For some, obtaining a consensus means getting support for their side. For them, the election process is a process of influencing opinions.

For others, obtaining a consensus means finding out whose side is supported by the majority. For them, the process is a statistical experiment, and influencing the result is the opposite of what they want to do.

For yet others, obtaining a consensus blends both of these concepts as a way for the members of the body of people to communicate with each other and come to a decision. This is not quite a middle-of-the-road approach, because it still leaves open the question of whether the winners have a responsibility to keep listening to the electorate or not.

Some ancient philosopher said, "Vox populii vox dei." And the Greek politicians tried to incorporate that concept in their government, in spite of the disagreement of many of their philosophers.

So, is the voice of the people the voice of God?

For those who consider the Book of Mormon to have some meaning, there are several salient verses. One reference is Mosiah 29:  26-27:

26 Now it is not common that the voice of the people desireth anything contrary to that which is right; but it is common for the lesser part of the people to desire that which is not right; therefore this shall ye observe and make it your law—to do your business by the voice of the people.

27 And if the time comes that the voice of the people doth choose iniquity, then is the time that the judgments of God will come upon you; yea, then is the time he will visit you with great destruction even as he has hitherto visited this land.

How is this different from vox populii vox dei?

One, it happens that the voice of the people chooses evil on occasion, especially when they are become so corrupt that there really isn't any help for them any more.

(Jonah, who had to have a whale return him to his duty, wanted Nineveh to be so corrupt, but even they weren't quite there yet.)

Two, with all the negatives, it is not common that the voice of the people desireth anything contrary to that which is right leaves a wide range of right possibilities for the people to be in favor of.

That the people usually don't choose the worst option does not mean that the people always choose the best.

You don't have to be a believer in the Book of Mormon to understand that much.

Why do I think this is important?

It demotivates the belief in the magic of winning, and the magic of winning is one of the enemies of freedom and of meaningful consensus politics.

With this emphasis on winning demotivated, perhaps it will be easier to understand the truth.

Statistics tell us that when sampling a population, a difference of a few percent is not likely to be meaningful. In general, 1 or 2% is near-equivalent to a tie.

A good statistical process will guard against biasing the result.

While that means that voter fraud should be guarded against, it would also mean that campaigning poisons the result. People change their votes based on all the argument and other exchange of rhetoric, and people are known to vote against their own opinions for a variety of reason.

Electoral processes have certain statistical natures, therefore 1 to 2% differences just are not meaningful. Ever.

But they are not, and, as long as we are dealing with humans, will never be proper statistical experiments. Therefore, even differences of 5% or more should be considered effective ties, at best.

Also, since any population should have some differences of opinion, large majority results are also indications of too much tampering with the sampling processes.

In plain English, 90% or more of the vote should be indication that something went wrong.

I'm going to stop there, and propose something really bizarre.

Our focus on who wins encourages fraud.

Can I repeat that? Heh. Well, go back and read that last one-sentence paragraph again.

If we had some way of recognizing a tie, we wouldn't be so concerned about things like the hanging chads in Florida in 2000. And if we had some way of reviewing wins by too large a margin, the majority would be demotivated in their attempts to make their majority position unassailable.

So I have a few proposals:

Any election where the difference between the top votes is less than 5% should automatically go back for a run-off. 

The first run-off should be subject to the same rule.

If, in the second run-off, the difference is less than 2%, the top vote getters should be considered to have tied. In the case of an office, the top vote-getters should share the office according to an agreement they work out together. The voters should have another ballot to approve the agreement, and if the ballot fails, the court competent to review the election should review the agreement and make recommendations.

In the case of regulations and laws, etc., the body competent to implement the regulation or law, etc., should review it, seeking a new regulation or law that will more effectively reflect the opinions of the various sides. Then the new regulation or law, etc., should be voted on again, with the same rules of effective tie. After the second effective tie, the competent court should review the question and determine whether further rewriting and election will produce useful results, and, if not, should have power to determine that the question has been rejected.

Any election where the top vote-getter gets more than 80% of the vote should be automatically reviewed for tampering by the court competent to do so. Opposition voices should be heard first and last in the review.

The court would have power to order a new election, if it determined that too much improper influence of any sort had been brought to bear during the election.

In any election where the top vote-getter gets more than 90% of the vote, the court should consider whether the top vote-getter should be disqualified from standing in the subsequent election for reconsideration. Reasons other than fraud may be considered, but the court must make its reasoning public.

In any election where the top vote-getter gets more than 95% of the vote, the top vote-getter should be automatically disqualified from standing in the subsequent election for reconsideration.

These kinds of rules would help to discourage the source of the problems that lead to voter fraud, and would also help to discourage the polarizing debates and voter hate that accompanies bad-faith electioneering.