Update on David G. Mills' Tennessee Lawsuit on the Unconstitionality of Pa

[Febble]

I filed my brief challenging the constitutionality of paperless voting in the Tennessee Court of Appeals on Monday, May 8, 2006. My lawsuit basically alleges that paper and paperless systems of voting, are substantially unequal systems and violate equal protection principles. My lawsuit aims to be the voting equivalent of Brown v. Board of Education.

My brief can now be found here in pdf format: http://dragonflihost.net/Mills_Election_Brief.pdf

Two hypothetical election contest lawsuits illustrate why there is so much inequality between systems.

The election contest lawsuit which demonstrates that paper ballots (even those which are scanned or printed on site) is a vastly better system to a paperless one is the case where the Court can have a jury count the ballots if the count by the election officials is extremely close or somehow in question. It would be a very easy and extremely verifiable election if a judge subpoenaed the ballots from the election commission, marked them as exhibits, and LET THE JURY COUNT them. Who would question that outcome? People who get to vote under a system where a jury could count their ballots if necessary, have an extremely good opportunity to have their vote counted accurately and would have genuine confidence in their vote.

Contrast that case with what would happen in a paperless election contest lawsuit. It would be a battle of experts between the challenger's computer software expert and the software experts of the election commission and all of the software experts from the voting manufacturer. That would be bad enough. But it could get far worse if the manufactures of the voting machines refused to give up access to their computer programs and even worse still if the manufacturer of the overall operating system refused to give up access to its proprietary software, and on and on with the manufacturers of every software driver of every peripheral, (modems, printers, etc.) not wanting to give up what they consider to be proprietary. Basically the challenger's expert may well end up knowing very little about the equipment and having to testify with both hands behind his back. Under this system, no voter would have confidence that his or her vote or the votes of others had been counted properly. This is what we have in my county.

Because the first kind of election contest is very simple and inexpensive, pre-printed paper ballots show why optiscan based systems can be verifiable. DRE's which produce a paper trail could also be verifiable but problems with printers and electrical shutdowns make them less ideal. It will take some auditing to have some idea as to whether it makes sense to have a jury count the ballots, but if there is a chance that a jury recount could change the outcome, it could be quickly, easily and inexpensively done.

Moreover, if there were too many ballots for one jury to count, that is not a problem. The courts can bifurcate trials and have two juries or as many bifurcations and juries as are necessary.

If you have any suggestions or comments about my brief please post them here or contact me at [e-mail]

[Febble]

1. Statement of the case

This is a voting rights case. The Appellant, a Shelby County attorney, alleged in both his Original Complaint and in his Amended Complaint that, as a citizen voter of Shelby County, he was being denied equal voting rights when compared to other citizens of Tennessee. It is his general allegation that citizens of other Tennessee counties are permitted to vote on paper or have paper records reflecting their votes while he and the other citizen voters of Shelby County do not vote on paper nor do they have a paper record of their vote. Appellant alleges that the inability to have a paper record of his vote diminishes his voting rights when compared with other citizens of Tennessee who get a paper ballot or record. Appellant alleges there is a disparity because without a paper ballot or paper record of his vote, there can be no meaningful recount in a questionable election outcome, nor can there be a meaningful election contest if one becomes necessary. Paper ballots or a paper record of the vote provide both.

I'm no lawyer, but it strikes me that this is a very interesting argument. Isn't it, in fact, a version of the argument that gave Florida to Bush in 2000? From Wikipedia:

Equal Protection Claim

Bush argued that the recounts in Florida violated the Equal Protection Clause of the 14th Amendment because there was no statewide standard that each county board could use to determine whether a given ballot was a legal vote. His argument was that since each county used its own standard to count each vote, some counties would have more liberal standards than other counties. Therefore, two voters could have marked their ballot in an identical manner, but one voter's ballot in one county would be counted while the other voter's ballot in a different county would be rejected, due to the varying standards.

[OnTheOtherHand]

2. yes, it is an equal protection argument

although Mills cites not the U.S. Constitution but the Tennessee Constitution.

As Mills points out in his brief, the bleeding-edge case law is in Stewart v. Blackwell, where a 3-judge panel decided (two votes to one) that using voting technologies that varied substantially in accuracy violated equal protection. (The word "substantially" is mine, I think not theirs, but the ruling does acknowledge some ambiguity about how large a difference would be consequential.) The majority opinion draws on Bush v. Gore, but it seems to me that the legal reasoning can hold even if Bush is ignored.

The finding in Stewart is "anti-punch," which is good, but it is also tacitly "pro-DRE," which is not so good. The courts just haven't worked out yet how equal protection applies to voting technologies.

[eomer]

3. There is a similar lawsuit pending in Florida.

Filed by Congressman Robert Wexler (D-FL):

http://wexler.house.gov/news.php?ID=218

-snip-

Wexler filed his federal lawsuit on March 8, 2004, under the precedent set by Bush v. Gore in the 2000 presidential election, which requires uniform standards throughout a state when conducting a manual recount. Because the 52 Florida counties with optical scan machines have a means to conduct a manual recount while the other 15 counties with touch-screen machines have a paperless system that cannot, Wexler is arguing that Florida is currently not in compliance with the Bush v. Gore standard and the equal protection clause of the U.S. Constitution. Wexler’s federal elections lawsuit is calling on the state of Florida to institute a paper trail in the 15 counties that currently use touch-screen voting machines. On December 7, 2005, the 11th District Court of Appeals heard oral arguments in Wexler’s federal elections lawsuit; the appeals court has not issued its decision.

[Febble]

4. It would be rather delicious

if Bush v. Gore was the ruling that got rid of paperless voting.

[eomer]

6. Turd blossom hoist by his own petard (french for "fart")

Word History: The French used pétard, “a loud discharge of intestinal gas,” for a kind of infernal engine for blasting through the gates of a city. “To be hoist by one's own petard,” a now proverbial phrase apparently originating with Shakespeare's Hamlet (around 1604) not long after the word entered English (around 1598), means “to blow oneself up with one's own bomb, be undone by one's own devices.” The French noun pet, “fart,” developed regularly from the Latin noun pditum, from the Indo-European root *pezd-, “fart.”

http://dictionary.reference.com/search?q=petard

[rumpel]

9. LOL

:rofl:

[Land Shark]

8. when you look at the definition of 'accuracy' used to distinguish

voting systems in Stewart v. Blackwell, a bit of wind will be let out of your sail.

Using the "residual vote" criterion as effectively the only way to look at accuracy is what Stewart v Blackwell did. (residual votes = overvotes plus undervotes)

David Mills would urge a broader view by focusing on other things that impact voting rights, which is great. But equal protection classes tend to get set up by statutory structures, and both HAVA and Ohio use this notice/non-notice dichotomy and "residual votes" flows from that and the EAC.

Although Blackwell set up the regulations for notice and non-notice, the plaintiffs then adopt those classes to their advantage.

The technologies are "blamed" in the end. Their is a lack of attention to causation of the problems complained of, contributed to in no small part by the fact that section 2 claims under the Voting Rights act don't need to prove causation, only disparate impact. But take a look at footnote six in Blackwell. The intra-county disparities in residual voting tend to prove that it is human decisions on allocation of technology, maintenance, training, respondind to troubleshooting, settings on equipment, and relative age of equipment that cause the various rates in residual votes that very by race, since it is all in the same county, it's all the same type of equipment!

The court devotes no attention to these issues and simply presumes that the varying rates are caused by voting technologies when in fact they are, at least in Summit County footnote six, most likely caused by decisions made by elections officials.

Note however, that it is clear that human agency can affect undervoting rates and perhaps overvoting rates as well. Consequently, savvy elections officials can manage their numbers in these areas through allocation of older machines and newer machines, maintenance, etc.

The worst possibility is that officials massage numbers to favor their favorite technology, whatever that may be. But it seems quite unlikely that paper will grow a mouth and warn us of undervotes, so I am still of the opinion that the statutory structures are a wee bit biased toward DREs.

Like I said in the other post below, David Mills and I agree that Stewart v Blackwell if not reversed en banc or by the Supreme Court will eventually require uniform voting systems statewide. We then have a very amicable parting of the ways as to strategies from there on out. But I think we'd agree that the full terrain is not quite clear. I don't yet have every fact I'd like to have and David agrees there are downsides to uniformity like making statewide fraud easier.

OTOH is too "sober" on the case, even the case itself admits that some see it as a "brave new world" while others see it as a twist on the old. To me, it's clearly a brave new world if pushing toward uniform standards and uniform systems is the direction (as it surely is) and if "non-notice" is going to be disfavored. That likely means the death knell for paper ballots if the courts do not broaden their ideas of "accuracy".

[Stevepol]

5. Thx Febble for this post. K&R!!

[Land Shark]

7. David Mills and I agree that Bush v Gore will "eventually REQUIRE...

"Eventually REQUIRE uniformity of voting systems" in a particular state.

The problem is, the way things are set up the tests for distinguishing one system from another used so far, such as that in stewart v. Blackwell (published just a few weeks ago) don't even allow discussion of our issues, and moreover the test itself is highly favorable to DREs.

The test: Add overvotes (for DREs this is always zero) to undervotes (DREs remind voters to vote undervoted races and not only will some voters be appropriately reminded, some more voters will think they HAVE to vote in order to cast their overall ballot, thus artificially depressing the DRE undervote rate). Together these are the "residual vote rate".

The test above is almost guaranteed to give DREs the best number, since it focuses, as does HAVA, on the one thing they are strong on which is overvotes/undervotes. In a pinch, and especially since the very survival of the technology is at stake in these kinds of "winner take all" equal protection analyses, the DRE manufacturer may have a single employee with a special needs child that really needs health care coverage currently provided by the manufacturer that will find it in their heart to make sure the program assigns most of the undervoted ballots to some candidate or another so that the DREs are guaranteed to have the lowest residual vote rate.

Thusly, we are guaranteed a 100% DRE country, over time. Unless the notice/non-notice dichotomy set up by HAVA and followed by the courts together with the residual vote test is rejected or modified to give a broader and more accurate view of accuracy....

As David Mills properly brings up in his lawsuit, there are many other considerations in a voting system comparison, such as recountability, etc., that are not being considered by the courts in favor of this narrow view of "accuracy"

[Febble]

10. some comments from David

OTOH : Although my suit is based upon the Tennessee Constitution and not the federal constitution, because Tennessee has never addressed the meaning of the words of our constitution which says that all elections must be "free and equal," I am asking the court to use the federal cases on equal protection as precedent.

This way I get to discuss and apply federal law to a state case. If the Court's rationale is good and if the Court discusses federal law in its opinion, then it will have some (perhaps a lot) precedential value in the federal courts as well as other state courts.

Also, I sure would like to get a copy of Wexler's brief. If the 11th circuit has not rendered an opinion by the time my case is heard, I would like to provide a copy of Wexler's brief to my court. If anyone can find a pdf of it, (I have not yet), please let me know.

Also both Land Shark and I do agree that Stewart v. Blackwell means we are headed for uniformity. We have had a long conversation about this by phone and he is more pessimistic than I, believing that uniformity will result in DRE's with no paper trail. He could be right.

But I think that is why shining a light on the farce that paperless voting makes out of the recount process and an election contest may prevent DRE's being cramdowned (to use Land Shark's term) our throats.

Consider this. The threat of a lawsuit (one with teeth and effectiveness anyway) deters wrongdoing. Lawsuits are also the final means of righting wrongs when everything else fails. I am a tort lawyer. In essence an election contest is basically a complaint by the Plaintiff that he has been the victim of a wrong ( a bad count, fraud, malfeasance) and he seeks to prove that he has been wronged and should have been declared the victor. If you make these cases of wrongdoing simple and verifiable, which you can do with paper ballots as evidence, then the court can right the wrong. If you make these cases extremely complex, a battle of experts, fights over trade secrets, and one where a decision is based on expert opinion not verifiable facts, then the election contest does not serve as a means of righting a wrong, as a means of deterring wrong, and becomes a legal fiction. Where is the due process if you have no means of proving you have been a victim of a wrong?

If we can work backwards, and begin by making sure we have valid election contests, which unquestionably verify a winner, (as paper would do if a jury could count the ballots), then I think we can convince courts we must have paper ballots or at least paper trails. Then those who would perpetrate fraud would know they might get caught and might not try to begin with.

But without the option to resort to a verifiable election contest, we have every reason to be pessimistic.

[eomer]

11. Febble, here are documents from the Wexler case (ones I could find)

Here is Wexler's brief for the appeal:
http://www.eff.org/Activism/E-voting/wexler/wexler_initial_brief.pdf

Here's where I found the brief (and there are a few other document links that may be useful):
http://www.eff.org/Activism/E-voting/

Another site that has some Wexler documents:
http://www.verifiedvoting.org/downloads/legal/florida/wexler-v-lepore/

[Land Shark]

12. We totally agree that if things don't change course in the law to a broad

or broader view, we are in trouble. David G Mills is attempting to do just that, broaden the court's mind. The Courts tend to resist mind-broadening experiences by saying they don't wanna be legislatures.

The real problem is that there ain't no legislature any where that has actually approved secret vote counting or the other issues we complain of. The election statutes prove that -- the entire structure is based on assumptions of paper and the legislatures have never wrapped their minds around the brave new world of electronics at any time.