Wexler: Please Don't take DAMAGED case to US Supreme Court

Opinion At: <http://www.ca11.uscourts.gov/opinions/ops/200416280.pdf>

CONCLUSION: Wexler v. Anderson again uses the misleading and damaging "residual vote" analysis also used in Stewart v. Blackwell (6th Cir. 2006) to uphold DREs, though this time in a slightly different different way, and involving the issue of meaningless DRE recounts. This case should not be sent up to the Supreme Court, IMHO, due to defective framing of the legal and political issues. You may skip to the bottom last two paragraphs if you don't want to the detailed discussion of why this decision ignores fundamental facts about voting systems.

DISCUSSION
It should be obvious to the average sixth grader that different voting machines are “gamed” differently, just like different video games are played differently. Thus, the way one might alter or cheat or malfunction on a lever system is different than the way one will alter or cheat or malfunction on an optical scan balloting system. But watch how the litigants in election litigation pay attorneys enormous sums to miss this basic point, as well as others. Our example du jour is Wexler v. Anderson, a published Eleventh Circuit case that came down June 20, 2006.

Following the Sixth Circuit in upholding DREs or issues related to DREs, the Eleventh Circuit in Wexler v. Anderson held that the manual recount procedures in Florida applicable to touchscreens do not violate the Fifth and Fourteenth Amendment equal protection and due process clauses of the United States Constitution. HOW it got there is the more interesting subject.

Background
A touchscreen recount in Florida (stupidly and irrationally) depends upon the percentage difference between the candidates, so therefore if you steal enough, you can avoid the following recount provisions.

(1) If the margin is within ½%, the manual recount consists of examining “The counters on the precinct tabulators to ensure that the total of the returns on the precinct tabulators equals the overall election return.” Fla. Stat. sec. 102.141(6)(a).



(2) If the margin is within ¼% or less, officials conduct a manual recount of all “overvotes” and undervotes” (Collectively, “Residual” votes). Fla. Stat. sec. 102.166. OVervotes are defined as voting for more than one on a ballot question, and undervotes note voting at all on a given ballot question. Fla Stat. sec. 97.021(23) and 97.021(37).

The purpose of the touchscreen recounts, as this august published case tells us, is to determine “if there is a clear indication on the ballot that the voter has made a definite choice.” Fla. Stat. sec. 102.166(5)(a). Now this is beyond dumb because a computer runs by 1s and 0s and doesn’t confuse the two, so there’s not going to be any appreciable chance of an ambiguous vote, but this does not by any means require us to accept or believe that the vote was recorded correctly. Standards for mechanical processes don’t translate into the different computer world paradigms.

After an administrative law judge struck down the “no recount” position in Florida, Florida then came up with the idea of printing ballot images and examining them for ambiguity, specifically as to their undervotes, to see if there is any “clear indication on the ballot image that the voter made a definite choice to undervote.” Rule 1S-2.031(4)(b)(2. The undervotes manually counted are then compared to the undervotes reported electronically. Id., at (b)5. If these two results which are guaranteed to match in fact DO match, the results are certified to the canvassing board. If they don’t match for some reason, the board must investigate “and resolve” the discrepancy. The emergency rule gets even more helpful, it says “clear indication” “shall be determined by the presence of the marking, or the absence of the marking, that the manufacturer of the voting system indicated shall be present or absent to signify an undervote.” Rule 1S-2.031(4)(a)2. Sequoia’s undervote, for example, is the absence of any information in the spot where a vote ought to be.

After a bench trial, the federal district court held that the Florida laws cited above established a “uniform nondiscriminatory” standard for manual recounts, even though the touch screen recounts differed substantially from the optical scan recount rules and methods, satisfying due process and equal protection.

The Eleventh Circuit held that the trial court was mistaken, in that compliance with Florida law did not define the correct issue (what the courts like to sometimes call “asking the right question”) The “correct question” is not whether these recount rules are “meaningful” as Wexler argued, but whether voters in touchscreen counties are “less likely to cast an effective vote than voters in optical scan counties?”

The “effective vote” standard is being boiled down to a residual vote comparison between systems. Paper based and lever systems create undervotes in malfunctioning and fraud-modes. Computers do not, or at least can avoid creating of undervotes.

The Court noted further in a footnote that issues of system “security, configuration and system malfunction are dealt with by the state during the certification process and are not relevant to the question before this court except to the extent that they make it less likely that voters using those systems will cast effective votes.”

The Court noted that the problem with the plaintiffs “nonmeaningful recount” approach is “that when a hypothetical touchscreen system “that is so nearly accurate that residual votes approach zero would nevertheless be constitutionally suspect if it was not susceptible to a substantially similar manual recount procedure to that applied in counties with far less accurate voting systems.” The Court also noted that the plaintiffs failed to plead that voters in touchscreen counties coast less effective ballots than those in non-touchscreen counties. The Court further held that a materially different type of recount is not so great a burden that strict scrutiny is appropriate.

Eventually the Court does note that touchscreens simply do not record ambiguous data about votes and that differing standards are ok for differing technologies that vary in the types of errors and mistakes that can occur. This is refreshing, but doesn’t have any effect on the Court’s analysis in terms of realizing that comparing residual votes for differing systems is not at all necessarily a “fair” test to determine the overall “accuracy” of a voting system. The Court also added that testimony by disabled voters added some weight to the state’s analysis, and prevent some of the errors voters can make on optical scan systems. There was no parallel discussion of the types of errors that voters make or systems make on touchscreens.

ANALYSIS
As with Stewart v Blackwell (6th Circuit, 2006) the “residual vote” analysis is unquestioningly asserted as the definition of “accuracy”. This microscope focuses the analysis on the one arguable strength of DREs and the “effective vote” concept implicitly puts the burden of proving “ineffectiveness” on the citizen under conditions were data establishing ineffectiveness with non-DREs is publicly available and data establishing ineffectiveness for DREs is largely kept secret under claims of trade secrecy, and subject to manipulation by vendors anxious to keep contracts and expand markets.

RECOMMENDATION
There are reports around the country that DREs are not letting people cast their vote until they vote all races (this will artificially depress residual votes, along with those who falsely conclude based only on a reminder by the DRE that they must vote every race in order to cast their ballot.) The residual vote analysis is a deck stacked heavily in favor of DREs. It’s hard to overestimate the danger posed by such a misleading legal standard.

This case will not fare differently in the US SUpreme Court. IT suffers from horribly damaged "framing", both legal and political. For all I know the plaintiffs did a great job briefing it, but the decision came out real bad, upholding DREs for the second time in two months in a federal circuit court.

Residual vote rates are another diversion.

They don't decide many races and there are other more significant problems with DREs. For one thing, you can't even be sure if the residual vote rate being reported by the DRE is correct! And they breakdown a LOT, making it impossible to vote. That's the best argument against DREs, IMO.

Florida has already made it illegal to hand count paper ballots, so the vote counting is secret throughout the state. There is no equal protection problem because NO voter in Florida is protected! Might be a due process argument though if you can have that without equal protection.

BTW, lever machines even prevent write-in overvotes by locking out the levers for candidate selections in the race(s) where the voter chooses to do a write in! That means if they had levers in FL in 2000, President Gore would be in the middle of his second term by now.

Jaws writes:

"There are reports around the country that DREs are not letting people cast their vote until they vote all races (this will artificially depress residual votes, along with those who falsely conclude based only on a reminder by the DRE that they must vote every race in order to cast their ballot.)"

Well maybe, but aren't residuals defined as top-of-the ticket over- or under- votes? It seems like your statement above is more concerned with down-ticket votes.

"The residual vote analysis is a deck stacked heavily in favor of DREs. It’s hard to overestimate the danger posed by such a misleading legal standard."

I do agree with this.

"This case will not fare differently in the US SUpreme Court. IT suffers from horribly damaged 'framing', both legal and political. For all I know the plaintiffs did a great job briefing it, but the decision came out real bad, upholding DREs for the second time in two months in a federal circuit court."

I agree. They must not be presenting the best arguments. Can they change them before appealing or must they appeal using the same arguments? (I'm not a lawyer you know!)

so the reported problems above would artificially depress undervotes, which is then used as evidence that DREs "save" votes, but in fact it would amount to forced voting, in the reported instances.

Raising arguments for the first time on appeal: It's rarely allowed and only in exceptional circumstances that new arguments can be raised on appeal. Usually only constitutional-level issues where there is "manifest error" or some obvious error on that level. That might possibly apply to this case, depending on the facts elicited at trial.

Anyway, what I'm saying is that residual vote statistics are only based on the top of the ticket, whether the DREs are forcing voters to make selections further donw the ballot or not.

However, if what you've reported is true, there may be other reasons for it, like forcing voters to cast votes in down-ticket races that cause vote switching to occur at the top of the ticket. Of course in some states, all it takes is one of those idiotic propositions to do that!